EDUARD BRUCKNER - THE SOURCES AND CONSEQUENCES OF
CLIMATE CHANGE AND CLIMATE VARIABILITY IN HISTORICAL TIMES
nico.stehr@zu.de
Eduard Bruckner - The Sources and
Consequences of Climate Change
and Climate Variability in
Historical Times
Edited by
Nico Stehr
Sustainable Research Development Institute,
University of British Columbia,
Vancouver, BC, Canada
and
Hans von Storch
Institute of Hydrophysics,
GKSS Research Centre,
Geesthacht, Germany
translations by
Barbara Stehr and Gordon Garnlin
Springer-Science+Business Media, B.V.
nico.stehr@zu.de
Library of Congress Cataloging-in-Publication Data
Bruckner, Eduard, 1862-1927.
Eduard Bruckner: the sources and consequences of climate change and climate
variability in historical times / edited by Nico Stehr and Hans von Storch.
p.cm.
Includes bibliographical references and index.
1. Climatic changes. 2. Paleoclimatology. 3. Brtlckner, Eduard, 1862-1927. I. Title:
Sources and consequences of climate change and climate variability in historical times.
II. Stehr, Nico. III. Storch, H. v. (Hans von), 1949- IV. Title.
QC981.8.C5 B75 1999
551.6'0903--dc21
99-058834
ISBN 978-90-481-5381-7 ISBN 978-94-015-9612-1 (eBook)
DOI 10.1007/978-94-015-9612-1
Printed on acid-free paper
All Rights Reserved
e 2000 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 2000.
Softcover reprint of the hardcover 18t edition 2000
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
nico.stehr@zu.de
Contents
About the Editors vii
Acknowledgements ix
Eduard Bruckner's Ideas - Relevant in His Time and Today
Nico Stehr and Hans von Storch
1. Introduction 1
1.1 Temporal Flow of Ideas and the Failure of Diffusion 1
2. Organization of This Book 3
3. The Climate Scientist Eduard Bruckner 6
3.1 The Life of Eduard Bruckner 6
3.2 Eduard Bruckner's Analysis of Climate Variability 8
4. Climate Change, Climate Policies and Society 11
4.1 Julius Hann and His View of Climate Variability 13
4.2 Climate Variability and Societal Importance 16
4.3 The Analogy to the Present State of Affairs 18
5. Conclusions 20
6. Bibliography 21
1. Groundwater and Typhus 25
2. Fluctuations of Water Levels in the Caspian Sea, the Black Sea, and
Baltic Sea Relative to Weather 47
2.1 The Annual Water Level Cycle 49
2.2 Secular Water Level Variations 52
3. How Constant is Today's Climate 63
4. Climate Change since 1700 77
4.1 The Current Status of the Inquiry into Climate 77
4.1.1 Climate of the Geological Past 79
4.1.2 Views and Opinions about Climate Change in Historical Times 88
4.1.3 Meteorological Cycles 116
4.2 Periodicity of Climatic Variations derived from Observations of
Ice Conditions on Rivers, the Date of Grape Harvest and the
Frequency of Severe Winters 127
4.2.1 Secular Variations of the River Ice 127
4.2.2 Secular Variations of the Time of the Grape Harvest 145
4.2.3 Secular Variations of the Frequency of Cold Winters 163
4.3 The Significance of Climatic Variations in Theory and Practice 171
5. About the Influence of Snow Cover on the Climate of the Alps 193
nico.stehr@zu.de
VI
6. Influence of Climate Variability on Harvest and Grain Prices
in Europe 219
7. Weather Prophets 243
8. An Inqniry about the 35-Year-Period Climatic Variations 255
8.1 Water Level Fluctuations in the Kirghiz Steppe and Fluctuations
of Rainfall in Russia since 1860 256
8.2 Decreasing Rainfall in the United States since the Middle of the
the 80's 260
8.3 Rainfall Fluctuations from 1830 to 1900 in the United States, as
well as at some Stations in Central Europe and East Siberia 262
9. About Climate Variability 269
10. Climate Variability and Mass Migration 285
11. The Settlement of the United States as Controlled by Climate
and Climate Oscillations 299
List of Publications of Eduard Bruckner 313
Climate 313
Glaciers 315
Glacial Ages 317
Morphology 319
Hydrology 321
Oceanography 322
Polar Research 322
Cartography 323
Biographical 324
Miscellaneous 324
Subject Index 327
Name Index 335
nico.stehr@zu.de
About the Editors
Nico Stehr is Senior Research Associate in the Sustainable Research Deve-
lopment Institute of the University of British Columbia, Vancouver, British
Columbia, Canada, and a visiting scientist at the Max-Planck Institut fUr
Meteorologie, Hamburg, Germany. He is a Fellow of the Royal Society of
Canada and editor of the Canadian Journal of Sociology. His research
interests center on the transformation of modem society into a knowledge
society, global change and public policy, the interrelation between climate
and society and the uses of social and natural science knowledge. Among his
recent publications are Practical Knowledge (1992), Knowledge Societies
(1994) and The Culture and Power of Knowledge: Inquiries into
Contemporary Societies (with Richard V. Ericson, 1992). His The Fragility
of Modern Societies is forthcoming.
From 1987 to 1995, Hans von Storch was Senior Scientist and leader of
the "Statistical Analysis and Modelling" group at the Max Planck-Institut for
Meteorology. In 1996, he became director of the Institute of Hydrophysics at
the GKSS Research Centre and professor at the Meteorological Department
of the University of Hamburg. He published several books, among others
"Statistical Analysis in Climate Research" (1999) and edited
"Anthropogenic Climate Change" (1999) and "Analysis of Climate
Variability" (1995).
His scientific interests are statistical analysis (especially transfer func-
tions relating large-scale climate to local features, identification of modal
structures in geophysical fields; data driven simulations), simulation of
regional climates and budgets of matter, paleoclimatic modelling, dynamics
and statistics of low-frequency climate variability, and transfer of knowledge
from natural sciences to the public arena (in cooperation with economists
and sociologists).
Vll
nico.stehr@zu.de
Acknowledgements
Before we acknowledge the crucial assistance and support of colleagues and
organizations we want to affirm our belief that the transmission of intellectu-
al traditions or the history of ideas and innovation in science are not simply
opposing activities. The transfer of ideas from the past should not be con-
flated with the notion that the preservation and acquaintance with knowledge
of the past equals mere repetition and preservation of these traditions. The
transmission of ideas from the past is always a mediation of such ideas in the
light of new circumstances and therefore present problems and issues. In
short, familiarity with past ideas can be instrumental in the construction of
new knowledge and not so much an obstacle to scientific discovery, as the
practice of science today often appears to imply. We would anticipate that
this also is the case with the formidable ideas Eduard Bruckner developed
about the dynamics of the climate system as well as its impact on society.
We are grateful to a number of individuals and institutions that have
made the English publication of this anthology of the wntings of the climate
scientist and geographer Eduard Bruckner possible. Gordon Gamlin has
provided us with a competent translation of Bruckner's writing. Barbara
Stehr has spent much imaginary energy and countless hours vetting and
improving upon the initial translation. We are most grateful to both. Skilful
and responsible editorial assistance was provided by Robin Taylor and Ilona
Liesner.
Some of the financial support in the form of research assistance was pro-
vided by the Canadian Social Sciences and Humanities Research Council.
The institutional support of Green College of the University of British
Columbia, Vancouver, Canada as well as the GKSS Forschungszentrum,
Geesthacht, Germany also was of considerable help in preparing this
volume.
IX
nico.stehr@zu.de
Eduard Bruckner's Ideas - Relevant in His Time and Today
Nico Stehr and Hans von Storch
1 INTRODUCTION
1.1 Temporal Flow of Ideas and the Failure of Diffusion
For a natural scientist, scientific discourse develops like the trunk of a tree.
Each year, a new tree ring is formed based on the most recent findings incor-
porating previous results-from the most recent tree ring, so to speak-and
newly established facts and interpretations. Knowledge obtained from earlier
research is either encoded or obliterated in present knowledge-continuous-
ly transferred from tree ring to tree ring-{)r forgotten. If something has not
been incorporated from cohort to cohort of scientists, it is considered to be
irrelevant and of little interest. This approach in natural science to its own
history is clearly manifest in practically all-contemporary articles in scienti-
fic journals. Most of the citations refer to work not older than 5 years. Some-
times casual reference is made to a handful of "classical" papers or books
but the authors have likely never closely examined these classic works but
know about them only indirectly.
This mode of operation is undoubtedly an efficient way of coping with
the sheer amount of pUblications scientists face daily. It is simply not possi-
ble to digest all new results-even in a field as relatively narrow as climate
science-let alone critically read many of the potentially relevant original
documents of past research. For example, for the process of understanding a
map displaying the global temperature by means of isotherms, it is not
important to know that the technique of isotherms was invented by
Alexander von Humboldt, or what his ideas about the technique were at the
time.
In almost all cases, this "diffusive"l transfer of knowledge from cohort to
cohort and generation to generation, or from "tree ring" to "tree ring", works
effectively and is robust enough to filter out what are considered to be irre-
levant constructions from the flow of knowledge. At the same time, during
I We use concepts geared towards the thinking of physicists. The relevant background is the
transport of heat in a fluid. Heat can be transported either by diffusion, which is maintained
by the collisions of the individual molecules within the fluid. This transport is relatively
inefficient as any transport is made up of many little steps over the small distances
between two molecules. In our metaphor, this means that knowledge is transferred through
personal contacts among scientists and through recent publications. A more efficient way
takes place when a current transports a parcel consisting of many molecules as a whole
over a longer distance. This process is called convective transport. In our context, this
refers to the introduction of forgotten concepts and results into contemporary thinking.
1
nico.stehr@zu.de
2 NICO STEHR AND HANS VON STORCH
this process of consensus building all knowledge claims are continually and
critically examined with respect to recent insights. Today, no one in the
natural science community would advance claims based on old authoritative
sources, as was common, for example, with the work of Aristotle during
medieval times. However, this process is not effective when a line of inquiry
in science is displaced for some reasons-and interest then reappears after a
longer period of time has passed and the collective memory about past
intellectual perspectives is no longer available in present-day journals and
scientists. In such a case, the transfer of knowledge needs more than
"diffusion" but outright "convective transport" from deeper placed tree rings
to the surface.
This "convective" transfer of ideas from the past should not be misunder-
stood as an attempt of merely repeating and preserving cognitive traditions.
Hand-in-hand with the transmission of ideas from the past goes a mediation
and interpretation of such ideas in the light of new circumstances and there-
fore present problems and issues. Thus, familiarity with past ideas can be
instrumental in the construction of new knowledge and is not so much an
obstacle to scientific discovery, as the practices of the scientific community
today often appears to imply but an intellectual asset in efforts to advance
SCIence.
We believe that climate science is a case for which the "convective"
influx of past ideas is a compelling necessity. After having been the book-
keeper for geography and meteorology in the 19th century, climatology
developed into a science of the physics and chemistry of the atmosphere and
the ocean; the early view that climatology is foremost a field of study that
deals with the impact of climate on people and society was virtually
forgotten. In the 1980s and 1990s, climate science underwent another para-
digmatic change: after the discovery that humankind is about to change
climate, the old problem of anthropogenic climatic change and the influence
of climate on individual and society re-emerged.
During our own work on the interrelations between climate and social
conduct we came across a number of early climate scientists who had a
significant impact on both their peers and the general pUblic. One of them
was the eminent geographer Eduard Bruckner, who is today forgotten in
climate science, and is considered by geographers to represent but a closed
episode in their disciplinary history.2 At the turn of the twentieth century, he
was one of the central protagonists in a vigorous debate in science and
society about global climate variability and its political and economic signi-
ficance. We believe that his formidable ideas could have a significant impact
on our present view of climate, climate variability and climate impact. It is
2 An informative overview about Bruckner's scientific career and achievements can be found
in Grosjean (1991).
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 3
for this reason that we have assembled this anthology of Bruckner's main
work on climate variability and climate impact.
2 ORGANIZATION OF THIS BOOK
In this introductory chapterJ we present information about Eduard Bruckner
and his scientific work, compare his approach with that of his contemporary
Julius von Hann, and relate his views to the present-day discussion.
The main part of this book consists of reprints of Bruckner's original
work in climate science. As most of his publications were in German, they
were translated. These translations of Bruckner's texts conform strictly to the
original. Only in the case of completely irrelevant notes have we decided to
delete such references. Additions we have made are inserted in square bra-
ckets. All diagrams have been redrawn. Some native city names used may be
less familiar than the English names: Miinchen is Munich; Wien, Vienna;
Praha, Prague.
The following is a list of the material presented. These eleven items were
chosen as they demonstrate well Bruckner's interest in climate variability,
his assessment of contemporary analyses and thinking about anthropogenic
climate change (such as the widespread concern about desiccation), and how
he has dealt with the transfer of knowledge into society.
1. Groundwater and Typhus [Grundwasser und Typhus. Mittheilungen der
Geographischen Gesellschaft in Hamburg], Volume III, 1887-1888.
2. Fluctuations of Water Levels in the Caspian Sea, the Black Sea, and the
Baltic Sea Relative to Weather [Die Schwankungen des Wasserstandes
im Kaspischen Meer, dem Schwarzen Meer und der Ostsee in ihrer
Beziehung zur Witterung] , Annalen der Hydrographie und Maritimen
Meteorologie, Volume II, 1888.
3. How Constant is Today's Climate? [In wie weit ist das heutige Klima
konstant?], Verhandlungen des VIII Deutschen Geographentages, 1889.
4. Climate Change Since 1700. [Klimaschwankungen seit 1700. Excerpts
from Klimaschwankungen seit 1700.] Wien; E.D. Holzel, 1890;
Chapter 1: The Current Status of the Inquiry into Climate Changes [Der
gegenwiirtige Stand der Frage nach den Klimaiinderungen.]
Chapter 8: Periodicity of Climatic Variations derived from observations
of ice conditions on rivers, the date of grape harvest and the frequency of
severe winters [Die Periodizitiit der Klimaschwankungen, abgeleitet auf
Grund der Beobachtungen iiber die Eisverhiiltnisse der Fliisse, iiber das
3 This introductory chapter incorporates some materials first published by Stehr et al. (1995)
nico.stehr@zu.de
4 NICO STEHR AND HANS VON STORCH
Datum der Weinernte und die Hiiufigkeit strenger Winter]
Chapter 9: The Significance of Climatic Variations in Theory and Practice
[Die Bedeutung der Klimaschwankungenfor Theorie und Praxis]
5. About the Injluence of Snow Cover on the Climate of the Alps [Uber den
EinjluJ3 der Schneedecke auf das Klima der Alpen], Zeitschrift des
Deutschen und Osterreichischen Alpenvereins, 1893.
6. Injluence of Climate Variability on Harvest and Grain Prices in Europe
[Der EinjluJ3 der Klimaschwankungen auf die Ernteertriige und
Getreidepreise in Europa). Geographische Zeitschrift, 1895.
7. Weather Prophets [Wetterpropheten], Jahresbericht der Berner Geogra-
phischen Gesellschaft, 1886.
8. An Inquiry About the 35-Year-Period Climatic Variations [Zur Frage der
35jiihrigen Klimaschwankungen] Petennann' s Mittheilungen, 1902.
9. About Climate Variability [ Uber Klimaschwankungen]. Mittheilungen
der Deutschen Landwirtschaftsgesellschaft, 1909.
10. Climate Variability and Mass Migration [Klimaschwankungen und
Volkerwanderungen). Talk at Kaiserliche Akademie der Wissenschaften,
Wien 1912.
11. The Settlement of the United States as Controlled by Climate and Clima-
tic Oscillations. Memorial Volume of Transcontinental Excursion of
1912 of the American Geographical Society of New York, 1915.
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 5
Eduard Bruckner and Albrecht Penck in the summer of 1893 on an excursion near Flims
(Graubunden, Switzerland). Taken from Budel, (1977).
nico.stehr@zu.de
6 NICO STEHR AND HANS VON STORCH
They say it is observed in the Low
Countries, that every five and thirty
years the same kind and suit of years
and weathers comes about again; as
great frosts, great wet, great drought,
warm winters, summers with little
heat, and the like, and they call it the
prime; it is a thing I do rather mention,
because, computing backwards, I have
found some concurrence.
Francis Bacon
3 THE CLIMATE SCIENTIST EDUARD BRUCKNER
3.1 The Life of Eduard Bruckner
Eduard Bruckner was born on July 29, 1863, in Jena, Germany.4 He lived for
a while in Odessa, Russia, before moving with his parents to Dorpat (now
Tartu, Estonia), where he spent most of his childhood. In [879 he was sent to
school in Karlsruhe, Germany. After graduating from high school, he studied
at the universities of Dorpat, Dresden, and Miinchen. He attended lectures
and seminars in geography, geology, paleontology, physics, meteorology,
and history. In 1885, he completed his doctorate under the supervision of
Albrecht Penck in Miinchen with a dissertation on the Glaciation of the
Salzach area (Die Vergletscherung des Salzachgebietes) in Austria. In 1886,
he moved to the Office for Marine Weather (Seewarte) in Hamburg to work
with Wladimir Koppen. The first two of our translated articles originate from
this early period of his scientific career. They pertain to the possible link
between groundwater levels and the incidence of typhUS, and the relationship
between sea water level variations and weather conditions.
On the strength of his dissertation, Bruckner was appointed professor of
geography at the University of Bern in 1988. He stayed in Bern for 16 years,
and became Rector of University of Bern in 189911900. During his stay in
Bern, he lectured on various aspects of geography but also regularly offered
public lectures. In 1904 he accept~d an offer from the University of Halle in
Germany and, in 1906, finally moved, as the successor of his former teacher
Albrecht Penck, to the University of Vienna. Bruckner died in Vienna in
4 Cf. Grosjean (1991) and Oberhummer (1927).
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 7
1927 at the age of 64. While in Vienna, he was, as in Bern, engaged in the
transfer of academic knowledge to the general public. He was chairman of a
series of Public University Lectures (Volksthumliche Universitiitskurse).
In 1890 he published the first extensive book-length discussion of recent
climate fluctuations, that is, of climatic fluctuations in "historical times".
Bruckner (1894: 1) credits the head of the Bavarian meteorological services,
C. Lang, with the discovery of decadal scale climate variability in a study of
the climate of the Alps. We have selected chapters 1, 8, and 9 of this
monograph (item 4).
After 1890, Bruckner published only a few smaller articles on the
observational evidence of climate variability (Bruckner, 1895, 1902; reprin-
ted as items 6 and 8). He explains the small number of articles on the
observational evidence as the result of a lack of new and appropriate meteo-
rological data on the issue. In the present-day context of particular impor-
tance, though, are his articles in which he speculated about the geographical
and socio-economic impact of climate change, i.e., the social consequences
arising from the climate fluctuations, such as emigration, immigration, and
migration patterns (Bruckner, 1912; [1912] 1915; items 10 and 11), or on
harvests, the balance of trade of countries and shifts in the political predomi-
nance of nations (Bruckner, 1894, 1895, 1909; the last two reprinted as items
6 and 9).
He was convinced that the issue of climate change and its impact was
both of considerable scientific merit and that future climate changes are of
great importance to the well-being of society as well as for the strategic and
economic balance of political and economic powers. He therefore presented
his conclusions about serious repercussions associated with climate change
anticipated for the end of the past century in the form of oral presentations
addressing the general public and especially affected segments of the public,
such as farmers. As a result, Bruckner presented his initial findings on
climate change not only to a congress of professional geographers in Berlin
in 1889 (our item 3), but also a year earlier in a public lecture entitled Is our
climate changing? at the University of Dorpat that was duly noted in the
local press (Bruckner, 1888). Later Bruckner (1894, 1909; the last one is
reproduced as item 9) published newspaper articles about the general issue
of climate change as well as about its specific economic and social conse-
quences. His work on climate variability was discussed at length in the
contemporary press (e.g., Neue Freie Presse, Vienna, February 11,1891).
As a result of these activities and the response they generated, Bruckner's
work on climate variability found a considerable echo among the scientific
community of climate researchers (e.g., DeCoumy Ward, [1908] 1918),
sociologists (e.g., Sorokin, 1928: 120-124), geographers (e.g., Huntington,
1915: 172-173; [1915] 1924:25), historians (e.g., Le Roy Ladurie, [1971]
nico.stehr@zu.de
8 NICO STEHR AND HANS VON STORCH
1988:217,220) and physicists (e.g., Arrhenius, 1903: 570-571), but to some
extent also among the public at large, as is exemplified by the fact that he
was often mentioned as an influential climate scientist in various
encyclopaedia until the 1950s. Huntington (1915:172) elevates Eduard
Bruckner to "one of the chief European authorities on climate" and credits
him for having initiated a kind of paradigm shift in climate research: "Since
the publication of Bruckner's widely known book on 'Climatic Changes
Since 1700' there has been a strong and growing tendency to treat climate
as a dynamic instead of a static geographical force" (Huntington,
1916:192).
3.2 Eduard Briickner's Analysis of Climate Variability
In the following section we summarise Bruckner's attempt to synthesize the
observational evidence for global-scale synchronous climate variability from
his limited data and limited computing power. Most of this synthesis is
described in his 1890 monograph.
Bruckner (1889:2) indicates that he was first alerted to the possibility of
climate change, aside from information about shrinking glaciers in the Alps,
as the result of observations about changing water levels in the Baltic, the
Caspian and the Black Sea (item 2). The changes in the water levels
appeared to follow a specific pattern. The rhythm of the changes resembled
changes in the glaciers of the Alps.
In his detailed discussion of "recent" climate fluctuations Bruckner
(1890) justified his approach by referring to the studies of E. Richter, C.
Lang and A. Swarowsky. Richter concluded that the causes for the secular
variations of one specific glacier (Obersulzbachgletscher in Austria) are wet
and dry periods lasting for several years in that particular region. Lang
showed that this result is valid for the entire Alpine region. Swarowsky
stated a striking correlation between the variation of the water level of the
Neusiedler See, a lake without any outlet near the Austrian-Hungarian
border, and the secular variations of the glaciers in the Alps, thereby demon-
strating that lakes without an outlet are excellent indicators of secular
climate variability.
In his 1890 monograph on climate variability, Bruckner started his analy-
sis with a careful investigation of the world-wide largest "lake" with no
outlet, the Caspian Sea. Bruckner drew the conclusion that Lang's results not
only hold for the Alps but may be extended to the vast catchment area of the
Caspian Sea (Bruckner 1890:86). He found that the climatic variation
followed a characteristic 35-year pattern, with wet and cool conditions
alternating with dry and warm conditions.
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 9
This inductive method of extending results from a smaller region to a
larger one is, by the way, typical for Bruckner's approach and consequently
he searches in data available from several other lakes without an outlet all
over the world for signals of secular variations. Bruckner states that the mere
existence of water variations in the lakes allows to the presumption that
secular climate fluctuations take place in the corresponding catchments
(Bruckner 1890: 115). In a further step, Bruckner applies the concept of
linking water levels of lakes to the rainfall in the corresponding regions also
to lakes with an outlet (Flusseen) and even rivers thereby stating the
existence of a more or less synchronous climate fluctuation over the entire
land mass of the world (Bruckner 1890:132).
The record of instrumental observations available to Bruckner reached
back for about 100 years. In these data he identified a rhythm of 35-year
alternating wet/cool and dry/warm episodes. In order to trace these charac-
teristic climate fluctuations further back, Bruckner also studied the observed
data of the ice conditions of the rivers, the grape harvest and the abundance
of strong winters. According to his data, Bruckner was able to identify 25
quasi-periodic cycles of about 35 years length during the last 1000 years
(Bruckner 1890:286).
He emphasized the fact that his mode of variability was not strictly
periodic but that the alternating wet and dry periods lasted about 35 years on
average. This fact is insofar noteworthy as in Bruckner's years, the fashion
of decomposing time series of all sorts into its Fourier components in an
attempt to describe the time series as a sum of predictable components
developed. Obviously Bruckner stayed away from this fashion, which later
was shown to be based on a simplistic misunderstanding of the mathematics
of statistical time series. 5
5 The fascination with the notion of periodic cycles as a description and an explanation for the
rise and fall of geological phenomena, of plants and animals as well as social and
economic processes, was still a vibrant enterprise in science during Bruckner's career. Sir
N. Shaw's Manual of Meteorology from the mid-1930s featured a list of several pages
length of various periods found in meteorological data. The conviction that "the whole
history of life is a record of cycles" (Huntington, 1945:453) was widespread. The
fascination arises from the fact that a process made up of a superposition of a finite
number of periodic sub-processes makes the process predictable: "It will be a vast boon to
mankind when we learn to prophesy the precise dates when cycles of various kinds ""Iill
reach definite stages" (Huntington, 1945 :458). In the 1920s and 30s, the Russian mathe-
matician Slutsky showed that the Fourier analysis of a statistical time series always reveals
some periodicities, even if the time series is constructed free of such periodicities. If
different chunks of such time series are analyzed, different periodicities pop up and vanish.
In spite of this finding, which today is completely understood, in some circles, and in
particular among lay scientists, the interest did not cease. On the contrary, in 1941 the
interest in the study of cycles led to the formation of a "Foundation for the Study of
nico.stehr@zu.de
10 NICO STEHR AND HANS VON STORCH
He speculated that the dynamical mechanism behind his quasi-oscillation
would be related to some unknown solar forcing mechanism (Briickner
1890: 240, 242) but was aware that no observational evidence for such an
oscillation exists. In this context Briickner denied any connection between
secular climate fluctuations and variations of sunspot activity (Briickner
1890:242).
Based on this 35-year period oscillation, Briickner prognosed a dry
period at the tum of the century (Briickner 1890:286,287) with severe
negative consequences in crops for continental regions, like Northern Ameri-
ca, Siberia and Australia. It is noteworthy that this predictive scheme would
have enabled Briickner to predict the "dust bowl" in the central part of the
United States, which actually took place during the Thirties of this century. 6
Briickner's methods were mainly limited to the exploratory statistical
analysis of time series since confirmatory tools such as confidence intervals
or hypothesis testing were not developed in combination with what might be
called common sense. He was unfamiliar with dynamically arguments (for
instance, concerning the geostrophic wind, which was well known among
meteorologists of those days) and he was unaware of theories concerning the
general circulation of the atmosphere (he failed to acknowledge the different
dynamic character of the tropics as opposed to the extratropical westerly
regime).
A fact that is impressive for modem climate researchers, who are used to
being supported by computers and digital data files, is the amount of
computational work done by Bruckner. It seems that he did all the calcula-
tions himself. He computed 5-year totals, called lustrum, and checked their
consistence by comparing records from neighboring stations. When data at
neighboring stations at some time begin to diverge, he concluded that one of
the two records is contaminated by some artificial effects, such as displace-
ment of the instrument (such as a water level meter). He tried to correct for
such inhomogeneities, and 'calculated correlations between his various time
Cycles" by Edward R. Dewey. The Foundation exists to this day and claims to have more
6 than 3000 members.
In 1915 Briickner predicted that by 1920 "we may expect a maximum of humidity" in the
United States (Briickner, 1915: 132). This prediction exploited two pieces of information:
first, the dynamical insight about the existence of a 35-year oscillation and second,
Briickner's finding that precipitation was at its minimum around 1900. On the continental
scale, his forecast was incorrect (Bradley, 1987: Fig. 6), but in a regional sense his
forecasts were consistent with actual developments: The Great Salt Lake exhibited
maximum lake levels from 1910 to 1930. Briickner did not spell out another prediction
based on the same reasoning, namely that in the middle of the 1930s the United States
would again suffer from dry conditions. Indeed the Great Salt Lake exhibited a sharp water
level drop in the early 1930s. Also the "Dust Bowl" dry episode that led to persistent
disastrous harvest failures in Central North America took place in the mid 1930s.
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 11
series to establish the degree of similarity between them. The sheer work of
just collecting the data, checking their consistency and calculating their
statistics must have been enormous, and hardly imaginable for a modern
scientist. His methodical approach is similar to what is done today when, for
instance, compiling records of the global mean temperature. The difference
is, of course, that the work is no longer done by human computers but
electronic hardware supervised by humans.
The number of hypotheses and theories
about climate change are numerous.
Quite naturally they have caught the
public attention, as any proof of past
climatic change points to the
possibility of future climate change,
which inevitably will have significant
implications for global economics.
Bruckner (1890:2)
4 CLIMATE CHANGE, CLIMATE POLICIES AND
SOCIETY
Today, the concepts of "climate variability", "climate change", and "climate
impact" attract an enormous interest not only in the climatological, meteoro-
logical, and oceanographic community (von Storch and Hasselmann, 1996)
but also in sciences concerned with climate-sensitive systems, such as
biometeorology, ecology, coastal defense, or the social sciences. The discus-
sion of "the climate problem"7 is by no means limited to the scientific
community. It has drawn a great deal of interest from a general public
(Lacey and Longmann, 1993) perhaps haunted by anticipations of catastro-
phic developments as a consequence of future anthropogenic climate change
(Stehr and von Storch, 1995). Evidence of a public and scientific preoccupa-
tion with "the climate problem" is given by such institutions as the "Intergo-
vernmental Panel on Climate Change" (lPCC) and international conferences
aimed at the establishment of International Climate Conventions.
7 We place the expression "climate problem" in quotation marks since it is not well-defined.
Natural scientists associate with this expression the understanding, prediction, and,
possibly, control of climate variability. Social scientists, on the other hand, consider the
perception of climate, and its social and political implications as the "climate problem".
nico.stehr@zu.de
12 NICO STEHR AND HANS VON STORCH
The majority of the scientific and general public interprets the climate
problem as a new challenge. Yet, although for much of past two centuries
most "climatologists" and meteorologists have been convinced, and have
considered it to be almost an axiom, that global climate is a constant during
historical times, 8 some 19th century climatologists, geographers and meteo-
rologists maintained that climate is not a steady phenomenon (e.g.,
Bruckner, 1890; Hann, [1883] 1893:362), recognizing that climate varies not
only on geological time scales (thousands of years and longer) but also on
decadal and century time scales due to natural and anthropogenic processes.
The processes that were discussed at the turn of the last century as the
source of climate variability and change were different. The "natural
variability", unrelated to man's activities, was speculatively attributed to
astronomical factors, such as the solar activity, and to processes in the
interior of the earth. In addition, the idea of deterministic periodic processes
attracted much attention among climatologists. Anthropogenic "climate
change" was thought to be the result of human activities, such as de- and
reforestation or new cultivation of land in North America. The possibility
that anthropogenic emissions of carbon dioxide might alter the global cli-
mate was first discussed by the chemist Svante Arrhenius (1896; 1903), but
dismissed by him as a realistic perspective for the next few hundred years.
The intensive debate among climatologists at the turn of the century
receded into the background when a new disciplinary consensus emerged
that remained predominant until recently, namely that the global climate
system contained overriding equilibrating processes providing resilience
against secular climate fluctuations; fluctuations that did occur were seen as
distributed around a fairly stable mean climatic condition. Any anomaly
extending for a few years would be canceled by an opposite anomaly at
another time. On average, nothing would change. One reason why the per-
ception of climate variations on historical time scales became unpopular may
be the rejection of "catastrophism" and the eventual acceptance of "unifor-
mitarism" in geology, as proposed by Lyell in the 1830s. Some of the social
8 Bruckner (1889:2) notes that during the 19th century, a distinct disciplinary division with
respect to the issue of climate change could be observed: Geographers and geologists were
more inclined to consider a persistent climate change to be a reality while meteorologists
defended the thesis that climate is a constant. Bruckner (1890:2) offers an explanation why
most professional meteorologists and many geographers at the time were rather silent on
the issue of climate change; as a matter of fact, he observes that they were embarrassed to
engage in research and discussion about climate change. The reason for the reluctance is
the wealth of competing hypotheses about climate change formulated earlier in the
century. But previous efforts only resulted in many contradictory voices about the nature
of climate change, so that climatologists then became reluctant to add to the cacophony of
mere opinions. Even in 1959, the prominent climatologist H. Lamb complains that many
of his contemporaries consider climate as something static. (Lamb, 1959)
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 13
scientific theories about the impact of climate on civilizations, for example,
by Sombart ([1911] 1951:324; 1938), Ploetz (1911), or Hellpach (1938), are
actually based on the explicit premise of constant climatic conditions (cf.
Stehr, 1996). In the public arena at the time, other urgent issues and concerns
displaced reflections about climate change and its impact on society.
In the following, we attempt to recover spirited discussions among geo-
graphers, meteorologists, and climatologists that occurred toward the end of
the last and at the beginning of this century. We make an effort to analyze
the dynamics of the discussion, and the degree to which it was introduced to
the general public, with the explicit intention of comparing the situation at
the time with the present discussions of climate variability and change and of
climate policies designed to avoid or mitigate the risk of climate change or to
allow for a smooth adaptation.
We concentrate on two of the main contributors to this early discussion
of climate variability and change on time scales of decades, namely the
already presented Eduard Bruckner and Julius Hann, both professors in
Vienna for a significant part of their lives. We will discuss their different
social roles, their attitudes towards the role of the public, and their under-
standing of their own work as part of multiple contexts in which they
attempted to play different functions. We will show that the two prota-
gonists, Bruckner and Hann, represent roles and self-conceptions that
resemble present-day roles of climatologists in discussions within and
outside the scientific community about the scientific significance and the
social impact of climate variability and change. We suggest that the "climate
problem", as perceived by scientists and the public at the turn of the century,
constitutes a valuable historical analog for present debates on the "climate
problem".
4.1 Julius Hann and His View of Climate Variability
Another leading and most influential professional climatologist at the turn of
the century was Bruckner's Viennese colleague Julius Hann, who was born
1839 in Wartberg, Austria. He studied mathematics, physics, geology, and
geography at the University of Vienna. After a career in teaching, he became
professor of physics at the University of Vienna and, in 1897, professor of
meteorology at the University of Graz. Between 1900 and 1910, he occupied
the newly created chair for cosmological physics at the University of Vienna
and served as director of the Institute for Meteorology and Geodynamics. He
died in 1921 at the age of 83 in Wien.
As Bruckner (1923:152) points out in his obituary, Hann may well have
been the most important meteorologist of his day and can be considered to
be one of the founders of modern meteorology as the science of the physics
nico.stehr@zu.de
14 NICO STEHR AND HANS VON STORCH
of the atmosphere (see also Steinhauser, 1951; Kahlig, 1993). He was des-
criptively oriented, that is, keen to establish the empirical or observational
basis for various meteorological phenomena. In meteorology, Hann disco-
vered, independently of Helmholtz, the thermodynamic theory of the Fohn.
In climatology and meteorology, he recognized early the importance of
quantitative methods and the significance of three-dimensional observation
systems, and he initiated the establishment of several mountain observato-
ries. In addition, Hann was editor of the internationally recognized Meteoro-
logische Zeitschrift for more than fifty years. Hann was an enemy of
speculative thinking; his main goal was to establish the facts (Bruckner,
1923:155).
Julius Hann compiled the first textbook on climatology. He first
published his Handbuch der Klimatologie in 1883. The Handbuch appeared
in a number of subsequent editions and translations and quickly became a
classic in meteorology and climatology (cf. Bruckner, 1923; Koppen,
1923:vi; Knoch, 1932:viii). An English edition based on the second edition
of the German version of the Handbook was published in 1903 (Hann,
1903).
In contrast to later editions of the Handbuch, its first edition summarizing
the state of knowledge in climatology then still defined as an auxiliary
science (Hilfswissenschaft) of geography (Hann, 1883:5; also Koppen,
1923:1) did not explicitly deal with the issue of climate variability. Reflec-
ting the preoccupation of the day with the issue of periodicity of climate,
Hann distinguishes between two types of climate fluctuations, namely "pro-
gressive" (that is, persistent transformations, or, in modern terms "climate
change"; e.g., von Storch and Hasselmann, 1996) and "cyclical" changes
(that is, fluctuations or oscillations around a constant mean with certain
characteristic times or periods; in modern terms "climate variability"). The
period for cyclical climate changes could be determined either by deductive
reason (by postulating a certain forcing mechanism, such as the sun's
activity) or inductive reasoning (by screening the observational record). It
should be possible, according to Hann, to trace progressive climate change to
either long-term trends of the temperature of the core of the earth or of the
output of the sun.
As far as relevant empirical material is concerned, Hann ([1883]
1897:390) referred to both non-instrumental and instrumental observations
of temperature and precipitation as well as general accounts or conclusions
about climate changes of a wide variety of observers found in disparate
historical records. He placed considerable emphasis on the critical examina-
tion of the observational climatic record. Obviously such data can be used
only if the procedure of observing, archiving and, possibly, correcting the
raw data is kept constant (cf. Jones, 1995). The historical data available to
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 15
Hann did in general not satisfy this homogeneity condition. He found on
close examination that the data recorded in the previous 150 years were
almost always contaminated by time-variable biases due to changing obser-
vational practices; the oldest instrumental records invariably were started in
rapidly expanding cities, and therefore reflect "urbanization", while rain
gauges were first placed on higher elevations (e.g., roofs) causing severe
biases in measurements (cf. Karl et aI., 1993).
On the basis of such methodical pitfalls concerning the quality of the
data, Hann was in general rather skeptical of scientific claims identifying
climate variability and change in the observational record. In particular, he
inferred that the evidence for systematic trends ("progressive changes") of
the climate during the historical period based on the available data from
different centuries, continents and countries is not substantial (e.g., Hann,
[1883] 1897:390). It had been hypothesized that the continental United
States of America of the 18th century was subject to an anthropogenic cli-
mate change due to the progressive anthropogenic transformation of nature
in the course the colonialization. Hann concluded with Whitney (1894), that
there is no hard evidence for a resulting climate change on the North
American continent (Hann, [1883] 1897:392).
In the case of climate variability Hann was less reluctant. He was
skeptical about strictly periodic climate fluctuations, especially in regard to
any hypothesized connection between variations in sunspot activities and
meteorological elements such as temperature, precipitation or changes in the
formation of glaciers. On the contrary, he concluded that the influence of sun
spot activity on climate patterns is insignificant. Moreover, he rejected the
possibility of any predetermination or causal linkage between climatic
variations and sun spot activities (Hann, [1883] 1897: 394).
Hann considered Briickner's quasi-oscillatory 35-year cycle much more
favorably since it was based on rich data from very different sources.
Briickner's discovery seemed valid for many regions and periods, and was
supported by many independent observations Hann ([1883] 1897:400) made
no serious independent attempts to clarify the dynamics of Briickner' s obser-
vational evidence. Instead he limited himself to efforts to establish the exis-
tence of the patterns of climatic fluctuations. Hann highlighted the fact that
Briickner's observations manage to shed light on contradictory accounts of
climate variations in specific localities since they "obviously" must have
been advanced during different phases ofthe 35-year period.
Indeed, the second edition of the Handbuch, published in 1897, contains
a forty-page separate section on climate variability that centers on
Briickner's research. In the fourth edition of the Handbuch, published in
1932, Karl Knoch had succeeded Hann as author of the Handbuch, (Hann
and Knoch, [1883] 1932). This fourth edition deals even more systematically
nico.stehr@zu.de
16 NICO STEHR AND HANS VON STORCH
with climate variability, even if the summary is rather skeptical. Much pro-
minence is given to contributions that attempt to demonstrate the stability of
the climate in historical times and point to the absence of evidence for
secular change (see also Berg, 1914).
4.2 Climate Variability and Societal Importance
It was and is common sense that climate variability and climate change have
a direct and powerful effect on many aspects of society, including the
economy, human health, or even the balance of power among nations. 9
Based on these views, the perspective of Climate Determinism emerged
suggesting that climatic conditions determine virtually all aspects of social
life, especially the chances of a society to attain a "high level of civiliza-
tion".10 This approach was widely accepted in geography and other disci-
plines at the time of Bruckner and Hann. 11 It is therefore of interest to inquire
how Bruckner and Hann responded to the challenge of offering their findings
to the scientific and general public as warnings of impending climate change
but also as instruments to design strategies to deal with climate variations.
Interestingly, the two scientists reacted very differently.
Hann disregarded societal impact entirely. He did not even mention
possible social consequences of climatic fluctuations. Consistent with the
then-prevailing self-conception of climatology as a largely descriptive (e.g.,
Hann and Knoch, [1883] 1932:3) "young" science (e.g., Koppen, 1923:v),
Hann examined the existing evidence on climate variability and change and
9 The impact of climate on the course of history has been of considerable interest. A more
recent account is Lamb's monograph Climate, History and the Modern World from 1982
(second edition 1995). Classical views have been put forward by ancient Greek authors
such as Hippocrates and philosophers of the enlightenment such as Montesquieu and
Herder. Also Friedrich Engels theorized about the influence of climate on society.
10 The perhaps most prominent representative of modem climate determinism is Ellsworth
Huntington (1915, 1945). For a discussion of the climatic determinism, see Stehr and von
Storch (1998).
11 One year after the publication of Bruckner's main work on climate variations, in 1891, a
certain Professor Umlauff published a scholarly textbook on the "Foundations of Meteoro-
logy and Climatology based on most recent research". In his introduction he claims that
"literature of different people is linked in a mysterious manner to the climate of their
homeland' ("So steht selbst die Literatur eines Volkes in geheimnisvollen
Zusammenhiingen mit den meteorologischen Elementen des von ihm bewohnten Theils des
Erdballes") and "Northern Europe has attained his superior level of civilization and moral
because of its rain throughout the year, whereas China's success in the past was related to
its summer precipitation." ("Nordeuropa habe es seinem Regen zu allen Jahreszeiten zu
verdanken, daft es der Sitz der hiichsten Gesittung wurde, so wie China seinem
Sommerregen die hohe Civilisation in fruher Zeit ... ")
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 17
attempted to establish whether the data supported arguments for changes in
climate phenomena.
Bruckner, on the other hand, not only discussed the nature and extent of
climatic fluctuations but emphasized their possible consequences for society.
In his 1890 monograph, he devoted an entire chapter to these matters: "The
importance of climate variability for theory and practice" (see selection 4 in
this book). In the terminology of present-day social science, Bruckner trans-
formed his academic findings into a form of "practical knowledge" (Stehr,
1991) that was meant to enable strategic responses in the economy, in the
field of transportation, health care and agriculture.
He argued that the area covered by ice fields varies, the size, the water
level, the appearance or even presence of lakes and rivers, the extent of
floods is sensitive to climatic variations. Such disturbances would have a
major impact on shipping and commercial patterns. Changing water levels
and the duration of ice covers on rivers and streams, in particular, would
affect the ability to navigate these waters and therefore the ease with which
goods may be moved. Another most important consequence would concern
agriculture more directly (see also Bruckner, 1894, 1895 and our selection 6)
since climatic fluctuations would have a significant influence even if the
effects depend to a considerable extent on the harvested product. Bruckner
concluded that more than two thirds of above-average agricultural outputs in
Europe with maritime climate coincided with the warm and dry periods and
an equal proportion of poor agricultural yields with the wet and cold climatic
periods. In more maritime climates enhanced summer rain would cause
harvests to be reduced whereas in continental climates, such as in Central
North America or Russia the summer rain would be favorable for agriculture
(Bruckner, 1894:2, 1915: 137-138).
Thus, the two phases of the 35-year cycle would both have a beneficial
effect in certain regions while disadvantages in other regions. Bruckner
(1915) concluded that as this specific pattern of agricultural productivity
change it would leave its marks on the temporal variations of emigration
from Western and central Europe to the United States. When conditions
were favorable in Europe, namely dry and warm, fewer people would
emigrate to the United States where a similar dryness and warmth reduced
the harvests. On the other hand, when cool and wet conditions prevailed on
both sides of the Atlantic, more people would travel across the Atlantic,
because agriculture in Europe suffered from the climate while productiv~ty
in the US was increasing.
Bruckner (1890:279-282) also proposed a connection between climatic
fluctuations and health. He dealt with one case in some detail, namely the
relationship between the appearance of typhus and the level of the ground
water, which is controlled by slowly varying precipitation amounts. Having
nico.stehr@zu.de
18 NICO STEHR AND HANS VON STORCH
examined records of typhus-mortalities in Central Europe, Briickner attribu-
ted at least part of the observed improvement in the mortality-in addition to
benefits derived from improvements in the sanitation system-to recovering
ground water levels as the result of a shift from dryer to wetter climates.
On the basis of his 35-year "mode of natural variability" and his analysis
of the climate sensitivity of civilization, Briickner (1890:279, 287; 1915: 132)
predicted a number of impending detrimental social consequences of clima-
tic variability, in particular serious economic crises for regions that had
benefited from a favorable climate in recent decades, especially areas located
within the continental climate regions, such as the United States, Russia and
Australia. These regions, Briickner argued, must expect an inevitable shift to
dryer weather resulting in significant crop failure.
4.3 The Analogy to the Present State of Affairs
The discussion among scientists at the tum of the century resulted in a series
of findings that present-day scientists would consider be sound and perhaps
of more recent origin:
1. Climate is not constant but varies on geological as well as historical time
scales.
2. Climate variability has to be differentiated between systematic, or, in
Hann's words, progressive changes and temporary variations, in Hann's
words, cyclical fluctuations.
3. The progressive changes were often related to human action (mainly
through land-use changes, often deforestation) while temporary fluctua-
tions were thought to be related to natural processes such a cosmic forc-
ings and processes in the interior of earth (including volcanic activity).
The main difference from today's discussion exists with respect to the
last point. At the tum of th~ century it was acknowledged that the dynamical
link between climate and extra-terrestrial variations was not firmly deter-
mined. Today, most scientists are convinced that a significant part of the
climate fluctuations has its origin in internal climatic processes related to the
non-linearity and stochasticity of the climate dynamics.
Scientists, then as today, were confronted with a number of scientific and
ethical questions. On the technical side, the problem arose of how to discri-
minate between human effects and internal processes? Should scientists
continue with the conventional curiosity driven research, or were the practi-
cal implications of observed patterns so serious that a purely academic
orientation should be given up in favor of a more applied research orienta-
tion? Because of the perceived importance of climate for political, economic
and social institutions, scientists were, and are today again, confronted with
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 19
the problem whether they should merely inform, or even warn society about
impending climate fluctuations and demand active intervention.
The protagonists of anthropogenic climate change, or, in modern terms,
environmentally conscious scientists did have an impact on the governmen-
tal-administrative level in different societies. Their message was that modifi-
cations of the environment were an agent of climate change (Grove, 1975).
For instance, the American Association for the Advancement of Sciences
(AAAS) in 1878 (quoted after Briickner, 1890:15) demanded reforestation
programs to avoid further desiccation that was perceived to take place in the
North America. Demands aimed at the abatement of anthropogenic climate
change were often met favorably by governments. In the 18th and 19th cen-
tury, governmental or parliamentary committees were instituted some coun-
tries, for instance in Prussia, Russia, France, and Italy (cf. Briickner, 1890:
14-19).
A further noteworthy fact that resonates the present predicament, is the
virtual silence of the "soft sciences" in the scientific and public debate. The
intellectual boundaries among scientific fields hindered the incorporation of
theoretical perspectives and empirical findings about climate that had been
advanced in other disciplines such as the evolving social sciences. That is,
the domains of the physical and the social milieu and their strict separation
had already become part of the social and intellectual structure of the scienti-
fic community.
Similarly, considerable energy was then, and is now being spent by
politicians, the public and others on the issue of climate change; in each
case, scientists played a major role in putting the issue of climate change
onto the public agenda. And politics endorsed the issue. In the past, how-
ever, the political response was mostly regional and not as it is now
demanded, global.
Briickner belonged to a small group of environmentally conscious and
socially responsible scientists. He and his colleagues felt obliged to inform
the public about the implications of his research. He was convinced that
climate varies for natural reasons. He considered the potential for predicting
these variations as most benevolent activity since it would allow govern-
ments and social institutions to anticipate and prepare for temporary
obstacles to social, political and economic developments. The task of the
scientist would be to first detect the regularities, and then to formulate and
convey the policy options arising from this predictive capability to govern-
ments and the public at large. Aside from informing the scientific communi-
ty about his results, Briickner does appear as far as we know to have not
addressed decisions makers directly. Instead he relied as indicated on
publishing newspaper articles and presenting public lectures. However, in
spite of his formidable insight into the climatological aspects, he, like some
nico.stehr@zu.de
20 NICO STEHR AND HANS VON STORCH
contemporaries today, overlooked that he did not have the expertise to
anticipate the societal response to pronouncements about pending adverse
climatic conditions, for instance by improving hygienic standards (the
typhoid forecast), by perfecting the railway system (the forecast concerning
the ice on the rivers) or by allowing for artificial watering of agricultural
land (the forecast concerning harvests).
Hann, on the other hand, remained an academic and restricted himself to
the immediate scientific problems at hand, that is, to the process of monitor-
ing climate and understanding meteorological processes. Why he refrained
from communicating more directly with the public and representatives of
different social institutions, we do not know. It could be that he did not
consider the results sufficiently firmly established, or he understood, unlike
Briickner, that leaving one's field of expertise creates a hazardous mixture of
scientific and political discourse that in the end may not be of any immediate
benefit to society.
In any event, the discussion and concern about climate change quickly
faded from the agenda in science and among the public in the first decades
of the 20th century. We an only speculate about the reasons. Certainly, some
of the practical promises associated with the new findings were found not to
be fulfilled, so that the whole story would be a case of "overselling", a
process later observed in the 1960s and 1970s with "rainmaking" and "cloud
seeding" (Cotton and Pie1ke, 1992). Also, the attention was diverted to more
pressing problems such as the big wars, the deep going social repercussions
and the economic disasters. Independently of the reasons, in the end a
consensus emerged among climatologists (e.g., Berg, 1914:67; Lamb, 1982)
that in "historical times" the global climate has been constant; that neither a
warming trend nor a trend toward less precipitation takes place. Moreover, in
climate science fascination with the results of new instrumental readings in
the 1920s and later shifted research attention away from the issue of climate
variability.
5 CONCLUSIONS
Our discussion of climate variability and climate change at the end of the
19th century leads to a number of conclusions that we consider relevant on
methodical, theoretical and practical grounds for present-day debates:
1. The discussion about natural climate variability and anthropogenic
climate change is not new. A similar debate, almost forgotten today, was
going on a century ag~. The protagonists found themselves in social roles
and situations similar to that of contemporary scientists.
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 21
2. The early debate on the nature and consequences of climate change
among climatologists, geographers and meteorologists lacked the interac-
tion with the philosophers and in the emerging social sciences who had
lively and vigorously written on the impact of climatic conditions on psy-
chological and social processes for decades and centuries.
3. The attention in the academia and in the public concerning the concept of
climate change and its societal implications was of limited duration. In
the end, the topic lost out in the competition with other economic, politi-
cal and everyday problems, and eventually also disappeared also from the
research agenda of the sciences.
The specific episode we have recounted reminds us that the burgeoning
genre of popularized science that surrounds present-day discussions of
climate change is by no means new. Nor is it novel to acknowledge the
uncertainties that surround scientific data on climate variability.12 It appears
that the issue of climate change lends itself well to such popularization.
Perhaps it does so because the issue goes to the heart of our modem common
sense understanding of the natural climate as benevolent and trustworthy (cf.
Stehr, 1997).
6 BIBLIOGRAPHY
Arrhenius, S. A., 1896: "On the influence of carbonic acid in the air upon the
temperature of the ground." Philosophical Magazine and Journal of Science
41:237-276.
Arrhenius, S. A., 1903: Lehrbuch der kosmischen Physik. Volume 2. Leipzig: S.
Hirzel.
Berg, L., 1914: "Das Problem der Klimaveranderung in geschichtlicher Zeit."
Geographische Abhandlungen 10 (2): 1-70.
Bradley, R. S., H. F. Diaz, J. K. Eischeid, P. D. Jones, P. M. Kelly, and C. M.
Goodess, 1987: "Precipitation fluctuations over Northern Hemisphere land
areas since the mid-19th century." Science 237: 171-175.
Bruckner, E., 1888: "Andert sich unser Klima?" Vortrag, Universitat Dorpat
(d. Neuen D6rptschen Zeitung No. 68)
Bruckner, E., 1889: "In wie weit ist das heutige Klima constant?", pp. 1-13 in
Verhandlungen des VIII. Deutschen Geographentages in Berlin. Leipzig: Teubner.
Bruckner, E., 1890: Klimaschwankungen seit 1700. Nebst Bemerkungen liber
die Klimaschwankungen der Diluvialzeit. Wien and Olmutz: Holzel.
nico.stehr@zu.de
22 NICO STEHR AND HANS VON STORCH
Bruckner, E., 1894: "Ru15lands Zukunft als Getreidelieferant." p. 1-3 in
Supplement to Miinchener Allgemeine Zeitung (November 19, 1894).
Bruckner, E., 1895: "Der EinflufS der Klimaschwankungen auf die
Ernteertrage und Getreidepreise in Europa." Geographische Zeitschrift 1:
39-51.
Bruckner, E., 1902: "Zur Frage der 35jahrigen Klimaschwankungen." Dr. A.
Petermann's Mittheilungen aus Justus Perthes' Geographischer Anstalt
48:173-178
Bruckner, E., 1909: "liber Klimaschwankungen." Mitteilungen der Deutschen
Landwirtschafts-Gesellschaft 24: 556-561
Bruckner, E., 1912: Klimaschwankungen und V6lkerwanderungen. Vortrag
gehalten in der feierlichen Sitzung der kaiserlichen Akademie der
Wissenschaften am 13. Mai 1912. Wien: K. K. Hof- und Staatsdruckerei.
Bruckner, E., [1912J 1915: "The settlement of the United States as controlled
by climate and climatic oscillations." Pp. 125-139 in Memorial Volume of the
Translantic Excursion of1912 of the American Geographical Society.
Bruckner, E., 1923: "Julius Hann." Pp. 151-160 in Akademie der
Wissenschaften in Wien, Almanach fiir das Jahr 1922. Wien: Holder-Pichler-
Tempsky.
Budel, J., 1977: Klima-Geomorphologie. Gebruder Borntrager
Berlin-Stuttgart, 304 pp. ISBN 3-443-01017-2
Cotton, W. R., and R. A. Pielke, 1992: Human Impacts on Weather and
Climate. Geophysical Science Series Volume 2. Ft. Collins, Colorado: ASTER
Press.
DeCourny Ward, R., [1908J 1918: Climate Considered Especially in Relation to
Man. New York and London: G.P. Putnam's Sons.
Grosjean, G., 1991: 100 Jahre Georaphisches Institut der Universitat Bern,
1886-1986. Jahrbuch der geographischen Gesellschaft von Bern, 56, 1986-90,
175pp.\
Grove, R. H, 1975: Green Imperialism. Expansion, Tropical Islands Edens and the
Origins of Environmentalism 1600-1860. Cambridge University Press
Hann, J., 1883: Handbuch der Klimatologie. Volume 1: Allgemeine Klimatolo-
gie. Stuttgart: J. Engelhorn.
Hann, J., 1903: Handbook of Climatology. Part I: General Climatology. New
York: Macmillan.
Hann, J. von and K. Knoch, [1883J 1932: Handbuch der Klimatologie. Fourth
Edition. Volume 1: Allgemeine Klimalehre. Stuttgart: J. Engelhorn.
nico.stehr@zu.de
EDUARD BRUCKNER'S IDEAS 23
Hellpach, W. H., 1938: "Kultur und Klima." p. 417-438 in Heinz Wolterek
(ed.), Klima-Welter-Mensch. Leipzig: Quelle & Meyer.
Huntington, E., 1915: "A neglected factor in race development." The Journal
of Race Development 6: 167-184
Huntington, E., [1915] 1924: Civilization and Climate. Third Edition, Revised
and Rewritten with Many New Chapters. New Haven: Yale University
Press.
Huntington, E., 1916: "Climatic variations and economic cycles." The Geogra-
phical Review 1:192-202.
Huntington, E., 1945: Mainsprings of Civilization. New York: John Wiley and
Sons.
Jones, P. D., 1995: "The Instrumental Data Record: Its Accuracy and Use in
Attempts to Identify the C02 Signal." p. 53-76, in H. von Storch and A.
Navarra (eds), Analysis of Climate Variability: Applications of Statistical
Techniques. Berlin: Springer Verlag
Kahlig, P., 1993: "Some aspects of Julius von Hann's contribution to modern
climatology." Interactions between Global Climate Subsystems. The Legacy of
Julius Hann. Geophysical Monograph 75.
Karl, T. R., R. G. Quayle and P.Y. Groisman, 1993: "Detecting climate
variations and change: New challenges for observing and data management
systems." J. Climate 6: 1481-1494.
Knoch, K., 1932: "Vorwort zur vierten Auflage." p. VIII-X in Julius von
Hann and Karl Knoch, Handbuch der Klimatologie. Fourth Edition. Volume 1:
Allgemeine Klimalehre. Stuttgart: J. Engelhorn.
Lacey, C. and D. Longrnann, 1993: "The press and public access to the
environment and development debate." The Sociological Review 41, 207-243.
Lamb, H. H., 1959: "Our changing climate, past and present." Weather 14:
299-318
Lamb, H. H., 1982: Climate, history and the modern world. Methuen & Co.
London
Le Roy Ladurie, E., [1971] 1988: Times of Feast, Times of Famine. A History of
Climate Since the Year 1000. New York: Farrar, Straus and Giroux.
Oberhummer, E., 1927: "Eduard Bruckner", pp. 195-199 in Akademie der
Wissenschaften in Wien, Almanach fUr das Jahr 1927. Wien: Holder-Pichler-
Tempsky
Ploetz, Al., 1911: "Die Begriffe Rasse und Gesellschaft und einige damit
zusammenhangende Probleme", pp. 111-136 in Verhandlungen des Ersten
Deutschen Soziologentages vom 13.-22. Oktober 1910 in Frankfurt am Main.
Tubingen: J. c. B. Mohr (Paul Siebeck).
nico.stehr@zu.de
24 NICO STEHR AND HANS VON STORCH
Sombart, W., [1911] 1951: The Jews and Modern Capitalism. Translated by M.
Epstein. Glencoe, Ill.: Free Press.
Sombart, W., 1938: Vom Menschen. Versuch einer geisteswissenschaftlichen
Anthropologie. Berlin: Buchholz & Weisswange.
Sorokin, P., 1928: Contemporary Sociological Theories. New York: Harper &
Brothers.
Stehr, N., 1997: "Trust and climate." Climate Research 8: 163-169, 1997
Stehr, N., 1996: "The ubiquity of nature: Climate and culture", Journal for the
History of the Behavioral Sciences 32: 151-159.
Stehr, N., 1991: Practical Knowledge. London: Sage.
Stehr, N. and H. von Storch., 1995: "The social construct of climate and
climate change." Climate Research 5:99-105.
Stehr, N. and H. von Storch, 1999: An anatomy of climate determinism. H.
Kaupen-Haas (Ed.): Wissenschaftlicher Rassismus - Analysen einer Kontinuitiit
in den Human- und Naturwissenschaften .. Campus- Verlag Frankfurt a.M. -
New York (1999), 451 pp ISBN 3-593-36228-7.
Steinhauser, F., 1951: "Julius Hann." In Osterreichische Akademie der
Wissenschaften (Hrsg.), Osterreichische Naturforscher und Techniker. Wien:
Gesellschaft fUr Natur und Technik.
von Storch, H. and K. Hasselmann, 1996: "Climate Variability and Change."
p. 33-58, in: G. Hempel (ed.): The Ocean and the Poles. Grand Challenges for
European Cooperation. Jena, Stuttgart, New York: Gustav Fischer Verlag
Whitney, J.D., 1894: "Brief Discussion of the question whether changes of
climate can be brought about by the agency of man, etc." United States
Supplement I, Boston. Appendix B: 290-317.
nico.stehr@zu.de
Chapter 1
Groundwater and Typhus·
The description of a disease according to its symptoms, its course and effect,
and the measures available to combat it, is the duty of a medical practitioner
and it would be presumptuous if a non-professional would venture to
participate in such matters. However, it is quite often the case that the
occurrence of certain diseases is caused by processes in nature and not in the
human body; in this regard those events are of particular concern, which are
the object of meteorological and geographical research. Illnesses are
triggered by certain meteorological events, e.g., the change between hot and
cold, dry and wet weather periods, though modem medicine has taught they
are not the actual cause, but contributing factors only. In the course of
investigating this correlation, while medicine moved into the territory of
meteorology, a new scientific area developed bordering on both,
meteorology and medicine, and being successfully researched by both. Its
borderline character should not be dismissed and certainly the line separating
both sciences should not be moved in favor of one or the other. This
situation may justify that the author, though being inexperienced in the area
of medicine, during a lecture at the Hamburg--Altona branch of the German
Meteorological Society on February 24th, 1888, turned to the subject of the
relation between groundwater and typhus, and in the following pages hands
his results over to the public, and at the same time would like to thank
Stadtphysikus Dr. Reincke for the unpublished statistical data of Hamburg's
typhus epidemic, offered so obligingly by him.
The incentive for the present study was the major typhus epidemic, which
invaded Hamburg between 1884 and 1887. In looking at the results of the
Hamburg medical statistics, one recognizes that beginning with the year
1838, the first year when related material became available, the number 'of
the annual deaths due to typhUS has decreased more or less continuously,
provided we exclude the very first years. Of 10,000 inhabitants, some 14 to
• Grundwasser und Typhus, Mitteilungen der Geographischen Gesellschaft in Hamburg, Vol.
III, 1887-1888.
25
nico.stehr@zu.de
26 EDUARD BRUCKNER
19 died each year in the first decade over this time span; by the end of the
Seventies and at the beginning of the Eighties, however, only 2 to 3 died.
Since 1885 those numbers worsened substantially. In 1885, of every 10,000
inhabitants in the State of Hamburg, 4 succumbed to typhus, in 1886, 7, and
in 1887, even 8 to 9, so that in the last year a total of 410 deaths were
registered as a result of typhus. How wide-spread the epidemic was is even
more evident if one not only looks at the number of deaths, but also at the
number of typhus cases: Of all inhabitants in 1879-84, an average of 700 fell
ill each year; however, in the 12 months from July of 1884 until June of
1885, this number totalled 1,334, in 1885/86 already, 3,015, and during
1886/87, even 5,330. The last figure is enormous because it indicates that
I % of the overall population was infected with typhUS at least once in these
12 months. Fortunately, the outbreak turned out to be fairly mild so that only
about 8-9% of those infected died in comparison with 15-16% in the years
prior to 1880. One question comes to mind immediately: Where is the cause
of this horrible calamity to be found?
His studies of the manner in which these epidemic or endemic diseases
spread led von Pettenkofer to the conclusion that certain factors in the
environment must be responsible for their occurrence, in that they enhance
or hinder the multiplication and spreading of the germs involved. All of
these external influences combined have been described by von Pettenkofer
as a local and temporal disposition. According to him, an epidemic outbreak
usually occurs in a particular location and only there if a number of
contributing factors is present, among which the meteorological ones playa
decisive role. In this way, he tried to explain the frequent and so strangely
localized outbreaks of cholera on the one hand, and pointed on the other
hand to a connection between the outbreak of typhus and the fluctuations of
the groundwater level. Most recently, I. Soyka in Praha has picked up on the
latter question, collected much new data and arrived at the same
conclusion.13 It is to Soyka's particular merit to have taken an experimental
approach in solving the questions about the role and influence of soil in the
development of germs. 14
Namely, the fluctuations of groundwater appear to be only contributing
factors, not the actual cause of the germs. The latter are tiny organisms,
fission-fungi, bacteria, bacilli, which when massively present in the human
body generate the dysfunctional and destructive symptoms, which add up to
13 Zur Epidemiologie und Klimatologie von Frankfurt a. M. [Epidemiology and climatology of
Frankfurt Main] Deutsche Vierteljahresschrift Itir Offentliche Gesundheitspflege XIX, 2.
Issue, Braunschweig 1887; Zur Aeti%gie des Abdomina/typhus, Archiv fUr Hygiene Vol.
VI, p. 257 ff.
14 Prager Medizinische Wochenschrift, 1885. Nos. 28-31, Zeitschrift fUr Hygiene Vol. II,
1887, p. 96 ff.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 27
what we describe as typhus. Such germs are omnipresent and everywhere,
they continuously penetrate the human body. Our body wages a constant
fight against them and tries to destroy these foreign invaders: if the body
does not succeed, the individual turns ill. However, there are situations and
certain conditions under which these germs multiply particularly fast not
within the human body but in the soil and for this reason occur in great
numbers, or some conditions under which they very easily come within close
proximity of humans. These conditions not only change from place to place,
but also over time and the number of germs invading the human body
changes accordingly. The immediate consequence is that at those times,
where the situation is favorable and fertile for those germs, the number of
illnesses grows and the disease takes on an epidemic proportion. A change in
these abetting or contributing factors or in their related causes results at the
same time in an immediate alteration in the outbreak of the disease. In this
sense, according to von Pettenkofer and Soyka, groundwater levels play the
role of secondary causes for the outbreak of typhus, the frequency of which
changes as fluctuations change.
Wherever solid rock is replaced by loose sediments, such as sand and grit
formations in the ground surface, we in our climate zone will find water at a
certain depth, which fills all hollow areas of the ground. It is infiltrated
rainwater, which collects on strata, such as compact rock or clay preventing
any further penetration, and which then accumulates as groundwater. Its
level rises or falls in time in the same way as the level of our rivers rises and
falls because on the one hand water is continuously added through rainfall
and on the other hand withdrawn through evaporation. That rainwater is the
source of groundwater has, however, been disputed recently by Volger. 15 It
can nevertheless be regarded as a proven fact. Soyka, after all, on the basis
of a large amount of data has only recently shown that the fluctuations of the
groundwater level do indeed follow the weather patterns on the surface. If
more rainwater reaches the ground level than is taken away through
evaporation, the groundwater will rise-in the reverse case, it will fall. This
is why the groundwater level fluctuates from month to month as well as from
year to year. A most striking fact is that the frequency of typhoid
occurrences follows exactly the same pattern as these changes of the
groundwater: whenever the groundwater drops or is at low levels, typhus
occurs particularly often; it is less frequent at increasing or high groundwater
levels.
Yet, at present, this opinion has still not found general acceptance in
medical circles, and it is safe to assume that more than half of all
practitioners remain convinced that not the fluctuations of the groundwater,
but impurities in the drinking water are the only cause of epidemic
15 Meteorologische Zeitschrift, 1887, p. 388.
nico.stehr@zu.de
28 EDUARD BROCKNER
outbreaks. Indeed, in some cases, as in the most recent typhus epidemic in
Zurich, the fact that the illness is limited to houses where drinking water is
supplied by one and the same water pipe seems to support this assumption.
Improvement of the drinking water is the key word whenever it becomes
necessary to fight back the typhus.
Finally, a third theory is added to these which focuses on soil movements
as an essential cause for an epidemic outbreak of typhus. If soil that contains
germs is turned over or dug up, these germs are brought into direct contact
with human beings initiating the outbreak of an epidemic.
Which of these perceptions may be considered the right one has not yet
been decided to this date, and it is quite possible that all three are correct
because nature is almost always in a position to achieve the same effect in
different ways. In my opinion, the typhus statistics of Hamburg in fact seem
to confirm this.
It was unavoidable that, during the previous typhus epidemic in
Hamburg, its cause was fiercely discussed. Almost unanimously, the
drinking water was found responsible for the disease, which admittedly is
bad enough in Hamburg: it consists of non-filtered Elbe River water with a
rich fauna thriving in the water pipe system. V. Pettenkofer's groundwater
theory was presented as well, by Dr. Reincke in particular, and finally many
voiced the opinion that the epidemic had been caused by the massive soil
movements undertaken to facilitate port and customs access. I will abstain
from giving a definitive answer to the question: however, I wish to present
some of the facts that could apply and may shed some light on a future
solution.
Let us first concentrate on the annual pattern of the groundwater levels
and the typhus frequency.
The annual fluctuation of the groundwater level depends on the inter-
connection of the annual rain season. and the intensity of evaporation during
the annual dry season. While the seasonal summer rains of central Europe
force the groundwater to reach its maximum level, evaporation being highest
during the summer strives to reduce the groundwater level to a minimum.
Depending on the location, wet or dry climatic conditions may be more
pronounced as they vary during the year. Yet the groundwater level adjusts
to either one or the other factor. Soyka discovered that in this regard there is
a contrast between Northern and Central Germany on the one hand and
Southern Germany on the other,16 the influence of rainfall prevailing in one
part of the country and evaporation in the other. Consequently, typhus cases
vary from month to month.
16 Soyka, Schwankungen des Grundwassers mit besonderer Beriicksichtigung der mitteleuro-
piiischen VerhiiJtnisse [Fluctuations of groundwater with special emphasis on central
European conditions], Penck's Geographische Abhandlungen, Vol. II, Issue 3, p. 49.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 29
Data for Miinchen are used to represent Southern Germany as shown in
the following chart. The monthly variation in evaporation is expressed as a
saturation deficit. The latter is the difference between the actually observed
absolute humidity of the air and the degree of moisture usually attributed to
the corresponding temperature. It is an indication for the ability of air to
absorb water and therefore an excellent measurement of evaporation. In this
case, it is arrived at by using the simpler though less exact method of
comparing the average temperature of a particular month and its
corresponding maximum level of humidity to that month's actual average
humidity reading. Figures for the groundwater level indicate its height above
sea level. The number of typhus deaths for each month is expressed in
percentages of the annual total. Throughout, the maximum is boldfaced, the
minimum is marked by an asterisk.
Miinchen ~1850-85}
Jan. Feb. Mar. Apr. May June
Rain 1 Precipitation mm 35 29* 48 56 78 112
Saturation Deficit mm 0.18* 0.42 0.86 1.84 2.43 3.11
Groundwater 51500 cm + 40 42 48 50 52 58
Typhus-deaths 17 % 11.5 11.9 11.2 9.0 7.5 6.9
July Aug. Sep. Oct. Nov. Dec.
Rain 1 Precipitation mm 112 102 72 54 50 46
Saturation Deficit mm 3.54 3.23 2.06 0.94 0.41 022
Groundwater 51500 cm + 59 57 45 37 32* 35
Typhus-deaths 17 % 6.4 6.5 6.3 5.8* 6.9 9.6
It is evident that on the one hand, the maxima of the groundwater level,
the rainfall, and the saturation deficit coincide and on the other hand, deaths
due to typhus decline to a minimum soon after the groundwater level has
peaked. In this case, it seems, the influence of evaporation is almost nil
because the effect of the rainfall is so much greater. In reality, however,
evaporation manifests itself in the fact that the groundwater level does not
increase as much in the summer as the annual period of heavy summer rains
might suggest if rain was the only influence. Due to the counteractive
influence of evaporation and rain, any fluctuation is quite low: the amplitude
amounts to only 27 em in Miinchen, while in other places where rain and
evaporation do not counteract to such an extent, the difference is much
greater, such as 58 em in Berlin with a more even rain dispersal, 60 em in
Bremen, 58 em in Brunn.
The groundwater conditions in the State of Michigan l8 are similar to
those in Miinchen. The average measurements of two locations showed the
17 Typhus 1856/85.
nico.stehr@zu.de
30 EDUARD BRUCKNER
following relation between rainfall, groundwater and typhus. The height of
the groundwater level is measured upward from a randomly determined
fixed point. I could not determine the saturation deficit for these locations.
MichiBian ~ 1885-862
Jan. Feb. March AEr. Ma~ June
Rainfall mm 73 31* 42 62 62 105
Groundwater cm 25 25 46 91 84 71
T~Ehus Mortali~ % 9 5 5 4* 4* 5
Jul~ AUBi· SeE· Oct. Nov. Dec.
Rainfall mm 62 128 116 51 67 54
Groundwater cm 51 20 33 20 8 0*
Typhus Mortality % 5 10 14 15 15 9
Here again the months of April to July show high levels of groundwater:
however, due to the increased evaporation during the summer, the absolute
maximum moves to April, while from July to August the water level drops
considerably and remains at a low level from August to February. In
complete accordance with this development, the incidence of typhus is
highest in October and lowest in April. The correlation between groundwater
levels [and precipitation] as well as typhus is shown in Figure 1.
Areas with less pronounced periods of rainfall show a very different
picture of the annual variation of groundwater levels. Berlin may serve as an
example for such an area.
Berlin {I 870-852
Jan. Feb. Mar. AEr. Ma~ June
Rainfall mm 40 35 47 32* 40 62
Saturation Deficit mm 0.71 0.95 1.55 2.73 3.95 5.13
Groundwater 3200 cm + 72 79 88 96 88 69
T~Ehus Mortali~ % 5.7 5.5 5.6 6.0 5.6 5.2*
July Au~. SeE· Oct. Nov. Dec.
Rainfall mm 66 60 41 58 44 46
Saturation Deficit mm 5.64 4.83 3.77 1.72 1.01 .59*
Groundwater 3200 cm + 56 45 40 38* 47 50
Typhus Mortality % 8.0 11.6 13.9 14.3 9.9 8.5
In this case, evaporation is indeed the predominant influence. The
groundwater level is highest in the winter months reaching a maximum in
April, a month with minimum precipitation; the level then drops with the
increase in evaporation although at the same time rainfall is increasing as
well. It starts to rise only after falling temperatures have reduced
evaporation. The typhus-scale again corresponds with v. Pettenkofer's
18 American Meteorological Journal, Vol. IV, p. 349 ff.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 31
findings: The months with high water levels show few cases of typhus, the
ones with low water levels show many cases.
Let us summarize our results in another way by calculating the annual
averages of the groundwater level as well as of the number of typhus cases,
as shown in the following chart.
Miinchen
Spring Summer Fall Winter
Groundwater in em 50 58 38* 39*
Typhus Mortality % 27.7 19.8* 19.0* 33.0
Michigan
SEring Summer Fall Winter
Groundwater in em 74 47 20* 17*
Typhus Mortality % 13* 20 44 23
Berlin
SEring Summer Fall Winter
Groundwater in em 91 57 42* 67
Typhus Mortality % 17.2* 24.8 38.1 19.7
Figures for Hamburg show the same pattern, as do those of all Northern
German cities, at least in the years before 1884. Admittedly, in the case of
Hamburg, the observations of the fluctuations of groundwater can only be
traced back to the year of 1880: they were carried out with great diligence by
the health official C. C. H. Muller in the suburb of Eimsbuttel not far from
the Isebeck canal, and who most obligingly submitted them to me. It does
little credit to the city of Hamburg, the second largest in Germany, that these
privately collected and funded data are the only ones available, while many
smaller cities have their wells checked continuously over a number of years,
such as in Bremen, 10 wells since 1869, in Frankfurt [a. M.], 15 since 1869,
of which 6 continue to be monitored, in Berlin, 31 since 1870, in Munchen,
1 well since 1856 with several others recently added, in Salzburg, 1 well
since 1860, and a larger number since 1867, etc.
Unfortunately, the location of the observation-point in Eimsbuttel is not
very favorable. The drill hole is not far from the local church, in the garden
of C. C. H. Muller, within a distance of about 100 m from the Isebeck canal.
The latter was dug in 1883 and 1884 replacing the Isebeck, a tiny brook
whose water level used to be some meters above the water level of the
present canal wide enough for the passage of larger boats. Inevitably, the
construction of the canal had some influence on the groundwater level of the
surrounding area. Therefore, when we observe that the groundwater level
fell by about 420 m from 1880 until 1887, we are certainly not mistaken, if
we allot most of the 273 m measured during the years 1884 to 1887 to the
nico.stehr@zu.de
32 EDUARD BRUCKNER
predominant effect of the new canal and only to a lesser degree to an overall
reduction of rainfall. Due to these factors, the observations cannot be used
for the time being as far as the annual averages are concerned. Meanwhile
the recently detected strong increase of groundwater levels in the course of
1888 suggests that those interventions by man did not tum out to be quite as
permanent or far reaching as it appeared. In any case, these observations
may serve well to demonstrate the annual fluctuations. With the exception of
course that the gradual drop of the groundwater level must have slightly
distorted the annual figures for the years 1884-87 in such a way that the
monthly averages at the end of the year appear lower compared with those at
the beginning of the year. This is indeed the case, if we compare tife
distribution of the figures for the annual averages of 1880-83 with the one of
1884-87. Both are quite similar, except the minimum [level] has shifted
from July, August, and September to August, September, and October. The
December level in 1880-83 was just as high as the January level, whereas in
the 1884-87 time frame it was considerably lower. Nonetheless, the overall
picture of the scale did not change and nothing keeps us from comparing the
annual distributions of typhus and groundwater.
Even a cursory glance at Hamburg's data on typhus and its annual
occurrence appeared to advise against expanding upon a common
denominator, but to suggest instead to consider the annual cases of 1880-83
and those of the epidemic cycle of 1884-87 separately and to compare the
result with the annual fluctuations of the groundwater during the same time
frames. The resulting chart shows the outcome. The typhus mortality per
month is expressed in percentages of the total annual number of cases.
Hamburg {1880-83}
Jan. Feb. Mar. AQr. May June
Groundwater in cm 17 19 19 15 11 5
TYQhus Cases % 7 6 5* 5 7 9
July Aug. SeQ. Oct. Nov. Dec.
Groundwater in cm 0* 2 3 4 9 17
T.YQh.us Cases % 10 15 13 9 8 6
Hamburg {1884-87}
Jan. Feb. Mar. AQr. May June
Groundwater in cm 22 25 24 25 20 12
TYQhus Illness % 12 8 6 4 3 3*
July Aug. SeQ. Oct. Nov. Dec.
Groundwater in cm 4 1 0* 1 4 II
Typhus Illness % 3 5 9 13 15 19
For the purpose of clarification, the numbers for Hamburg and also those
for Michigan are graphically depicted. In the scale system chosen, an
increase by one division equals an overall increase of 2% of the typhus cases
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 33
as well as an increase of 3 cm of the groundwater in Hamburg, or 15 cm in
Michigan.
r'" "- .* ..
.·
J .* .....
~
.
.··
.. / I\. 1/ I\. "
,, ,,
'I
MiobigM { ,1/ \.. ~/ 1\
J~
'/ ... .. . ..
:10' I"
.... If
~
.. -. .'
.. :./
'"
-
~
...., ·,,
.... ~ ~
1\ ~
I' \. , J
\
.
I
, . : ',
,,
I
.
'1880183 { If I
.* .. ... . ~
.",
.••• * -.
.. .'
.'
\
':j
...
Hamburg '1/ '. """
....
......
.... L..o
.
L..ooo
I' 1\ "' 1\ .
V V'
: ••J
. 'i1/' " .. ..
",J
1884187 { !/
,,
1\
. \
..· 1/J '... 1\\ ..·
, ,
1\ ~
.. . .....
,
,,
. •.•. \
. .. ....
171
,
'"
"'""~I/ 1.,..00 I
DJFMAMJJ ASONDJFMAMJJ ASOND
- - - - - Annual cycle of groundwater
••••••••••••• Annual cycle of typhus
Figure 1.1. Annual Twin Curve Groundwater (solid) and Annual Twin-Curve Typhus
(dashed)
At first glance it is evident that for the years 1880-83, the Hamburg
figures comply indeed with v. Pettenkofer's and Soyka's findings. The
number of typhus cases is highest at low groundwater levels and lowest at
high levels. With rising groundwater levels, this number shows a steady
decline. For the years 1884-87, the Hamburg figures show an entirely
different picture. Thus, contrary to what one would expect according to
v. Pettenkofer's theory and the observed tendency of the years 1880-83, the
opposite occurs. This contradiction is even more pronounced if we consider
that our annual curve is somewhat distorted by generally low levels of
groundwater. If we discount this distortion, we notice that the minimum
groundwater level moves even closer to the minimum number of typhus
cases, and the maximum of typhus occurrences moves closer to the time
when groundwater levels are high.
nico.stehr@zu.de
34 EDUARD BRUCKNER
These fluctuations of the groundwater level show the same pattern as
those that occurred in 1880-83, except for a minor delay in their onsets. In
Hamburg, prior to 1884, typhus used to be a summer epidemic peaking in
the months of July, August, and September. Since 1884 it has become a
winter illness with a maximum number of cases in November, December
and January. In order to visualize this more clearly we average the number
of typhus cases per season.
Hamburg: Frequency of Typhus Illness in %
Spring Summer Fall Winter
1880-83 17* 34 30 19
1884--87 13 11 * 37 39
The outbreak of the epidemic changed the seasonal typhus occurrences
completely and disrupted the previously clearly defined correlation between
groundwater level and typhus. This may well indicate that this unusual
epidemic had nothing whatsoever to do with the fluctuations of the
groundwater level, but rather with another unknown cause.
Let us now tum to the question whether, in a year-by-year comparison,
the change in the number of typhus cases may be connected with
fluctuations of Hamburg's groundwater. Soyka has investigated this for a
number of cities, such as Berlin, Frankfurt a. M., Bremen, Miinchen, and
Salzburg. He came to the conclusion that those years with major typhoid
epidemics also showed low groundwater levels and on the other hand years
with high groundwater levels appeared relatively typhus-free. To illustrate
his findings, we repeat the data for Miinchen. The number of typhus deaths
relates to 10,000 residents and the groundwater level is measured in
centimeters, beginning at 514 m above sea level. The data for the
groundwater refer to the Karlstrasse well, however, since 1875 to the one at
the [Miinchen] Physiological Institute. These data were adjusted by +6 mm
to the sea level of the Karlstrasse well based on the measurements for both
wells of the years 1869-74. The maximum is always boldfaced and the
minimum marked by an asterisk (*).
The years 1858, 1864, 1872, and 1879 stand out both for a high number
of typhus deaths and low groundwater levels: the years of 1861, 1867, 1876,
and 1878 for high groundwater levels and an extraordinary decline of typhus
cases. Even under more detailed examination a certain inter-correlation is
maintained; if groundwater rises from one year to the next, the typhus cases
decline at the same time. Of a total of 25 cases prior to 1881, 19 follow this
norm and only 6 are exceptions. This confirms the pattern that evolves on
the basis of the annual averages of both factors. Conformity is lacking from
1882 on due to reasons discussed later.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 35
Miinchen: T):]~hus Mortali!):: and Groundwater
1856 57 58 59 60 61 62 63 64 65
Typhus 29.1 28.2 33.0 17.0 10.7* 11.9 20.1 16.1* 25.3 20.5
Ground- 103 91 88* 123 140 152 131 120 125 105*
water
1866 67 68 69 70 71 72 73 74 75
Typhus 20.8 5.2* 7.4 11.6 15.0 13.3* 23.3 12.7* 16.0 11.7
Ground- 113 180 161 135 125* 137 127* 132 115* 120
water
1876 77 78 79 80 81 82 83 84 85
Typhus 6.0* 8.0 5.1* 10.4 6.4 1.8 1.7 1.9 I.5 1.7
Ground- 186 180 181 158* 175 179 135 141 112 112
water
Unfortunately, as mentioned before, only since 1880 do we have
groundwater levels recorded for Hamburg and even these are inconsistent.
Consequently, a direct comparison between the variation of typhus
occurrences and of groundwater is not possible. However, we are helped in
this predicament by an idea that can be backed up with numbers. The curve
of the groundwater level measurements can be replaced with the curve of the
river levels without falsifying the outcome. At first I believed this goal was
reachable in a different way. Namely, there are two factors that determine
the level of the groundwater: in addition to rainfall, which increases the
groundwater, we need to consider evaporation, which continuously absorbs
the groundwater-as previously mentioned. Only the interaction of both
factors determines the groundwater level and its fluctuations. That is why,
seen annually, depending on the predominant influence of the one or the
other factor, the groundwater level was at times reacting more to changes in
rainfall and at times to those in evaporation. For a year-long prognosis, it is
therefore impossible to determine a priori whether in a specific location the
groundwater level will adjust to the one or the other. The development over
several years is a somewhat different matter. Rainy years tend to be both
moist and cool, with little evaporation; whereas in very dry years
evaporation is high. The following five-year averages for Salzburg seem to
prove this correlation.
Groundwater Level Quantity of Rain Saturation Deficit.
Above Sea Level
41Om+ mm mm
1861165 2.97* 1058* 2.38
1866/77 3.03 1207 1.95
1871175 2.99 1283 1.70
1876/80 3.13 1431 1.63*
1881/85 3.04 1285 1.79
nico.stehr@zu.de
36 EDUARD BRUCKNER
The precipitation minimum coincides with the evaporation maximum and
vice versa. What applies to five-year averages does, to a certain extent, also
apply to each individual year. If for Salzburg individual years are ranked
according to the amount of precipitation and then bundled in a group of five,
and if, finally, the saturation deficit is determined, it is evident that the
following figures correlate with each other:
Average Rainfall Average Saturation Deficit
Quantity
inmm inmm
1546 1544
1361 1752
1278 1866
1111 1994
866 2280
Finally, if one looks at the changes in precipitation and in the saturation
deficit on a year-to-year basis, an increase in precipitation is most often
accompanied by a decrease in evaporation and vice versa. The figures for
Miinchen, for instance, show this reverse relationship for the years 1860 to
1885 in 17 cases, while in only 7 cases the two factors are changing in the
same direction; for Salzburg the figures are 15 and 7 respectively. Thus, rain
and evaporation have an equally strong effect on groundwater levels.
Indeed, it is Soyka's opinion as well that, on a larger scale, the variations
in precipitation and groundwater level are the same from year to year. Lang
even goes so far as to attempt to replace groundwater fluctuation with total
amount of precipitation. 19 Nevertheless, in compiling groundwater and
precipitation figures for Berlin, Frankfurt a. M., and Bremen, I became
convinced that such an approximation can be flawed as far as individual
cases are concerned, and, therefore, does not serve our purpose. I had to look
for some other entity and the water levels of the Elbe River seemed what I
was looking for.
Soyka pointed oufo that river levels and groundwater levels of
neighboring areas show a similar distribution pattern; he proves this
similarity with data for Berlin (Spree River), Frankfurt (Main River),
Bremen (Weser River), and Miinchen (Wiirm river).21 How strikingly
parallel the distribution lines for the two entities are is demonstrated in the
19 Schwankungen der Niederschlagsmengen und Grundwasserstiinde in Munchen 1857-1886
[Fluctuations of the Amount of Rainfall and the Groundwater Levels in Munich
1857-1886], Beobachtungen der meteorologischen Stationen im Konigreich Bayem, Vol.
IX, 1887, p. XIII.
20 Schwankungen des Grundwassers. etc., [Fluctuations of Groundwater, etc.], p. 80.
21 The Isar river cannot be used here because the depth of its river-bed changes rapidly.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 37
graphic shown above, which was drafted on the basis of Soyka's data. In this
graphic, one grid length corresponds to 0.1 m groundwater level change for
Berlin, 0.17 m for Bremen, 0.2 m for Munchen, and to 0.1 m river level
change for the Spree, 0.2 m for the Weser and 0.05 m for the Wurm.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I- I I I I I I I I I
. ..
............ groundwater Berlin
• f'.
. . .. 1870-85
.....
I- river level 1-•
1\ ... • r--
1 .' V 1"\' groundwater
.
1 jr- ~ Spree
. . ..
, fj r\.
Bremen ~
., ;;;;"'-
. 11869-84 ~1\
r'\: \
'" ,... 1 .
:~
"
,...
groundwater
Miinche; ~ .' Weser
.... .' ..
1856-85
I \.
,'. ., .,....
II •
."1 .,-
"
J I \ ~ 1/ I\. /1\.
. .. I
L .'
'.
'\
J.
. '
..~-I
J
•
7
J. \11
. Wiirm
., groundwater
~I"" II
/ '\ ~
"
Figure 1.2. Curve a/Groundwater Levels (dashed) and Curve a/River Water Levels (solid)
This chart gives strong evidence that fluctuations of the groundwater
parallel those of recorded river levels and that the one can be used instead of
the other without introducing any major distortion. This holds true for long-
term changes in water levels as well as for short-term. Groundwater and
river levels fall and rise in tandem. The figures for Berlin show changes in
the same direction in 11 cases and only 4 in the opposite direction, for
Bremen, 13 and 2, and for Munchen, 24 and 4, i.e., of these a total of 48
cases show similarity, and only 11 the opposite. This conformity is even
more convincing, when one considers that in only 1 of those 11 cases (near
Munchen in 1860-61) the fluctuation is of a larger magnitUde. We should
not be too surprised though because the amount and distribution of seasonal
precipitation and evaporation affect the water level in the ground and in the
river in exactly the same way. In addition, in places like Berlin and Bremen,
the groundwater interacts with the river water up to a certain degree and
consequently follows its movements,22 except, of course, in the case of
Munchen, where the groundwater obviously does not fluctuate or
interconnect with the distant Wurm River or even with the Isar River, a fact
that does not invalidate the observed congruity entirely. In this case it must
22 Soyka op.cit. p. 58, 80.
nico.stehr@zu.de
38 EDUARD BRUCKNER
be the equally strong influence of rainfall and evaporation alone that causes
confonnity.
Let us examine the relationship of groundwater and rivers in Hamburg.
The Elbe River penetrates the central and southeastern parts of the city by
way of numerous canals; the entire northern part of the city, however, is
under the influence of the Alster River and its tributaries, such as the
Isebeck, etc. The ground in Hamburg consists of either clay or sand
occurring in a fairly irregular pattern. In many places, clay boulders have
fonned ridges up to above the water level with basins in between or detached
knolls. Quite often the ground is soft sand down to below the riverbed level.
It is more than likely that under such conditions in Hamburg, as in Berlin,
Bremen and Frankfurt, groundwater and river water levels are closely
interrelated. Regrettably though, I cannot prove this relationship since
measurements of the Elbe River levels near Hamburg are available to me up
to the year 1869 only, whereas groundwater level measurements did not start
before 1880 and were restricted to the Alster area. The nature of the seasonal
changes, however, may be alluded to: groundwater and river water levels
reach their peaks in February, March, and April and their lowest levels in
September. But in view of the conditions found in Bremen, Munchen,
Berlin, and Frankfurt a. M., I have no hesitation in using the curve of the
fluctuations of the Elbe River levels in place of the curve of the groundwater
levels in Hamburg.
The following chart shows typhus deaths per 10,000 residents of
Hamburg for the years 1838 to 1887: furthennore, data on typhus cases for
the years 1872-87 are included. For this infonnation, as for all the other
infonnation on typhus statistics in Hamburg, I am indebted to Dr. Reincke,
the city's health official. Unfortunately, the annual averages do not refer to
the calendar year but relate to the period from July 1 to June 30 of the
following year under which they are listed respectively. On the other hand
the table includes the average water levels of the Elbe River for three
different locations: Hamburg, Artlenburg about 45 km further upstream, and
Magdeburg about 280 km upstream. The numbers for Magdeburg and
Artlenburg were arrived at by using a table published by Bureau der
Baudeputation Sektion for Strom- und Hafenbau. The locations Artlenburg
and Magdeburg were added because the water levels of the Elbe at Hamburg
are not merely the result of climatic conditions upstream. It is the tide that
influences the water level in the city of Hamburg to a considerable degree.
High tide, for example, averages 1.8 m annually which is four times the
amount between the highest and lowest annual mean water level. In addition,
the submitted annual averages are not the result of actual measurements, but
are arrived at by using the data on the averages of high and low tide.
Therefore, these data carry a flaw that was introduced by those tidal ranges.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 39
The charted water levels are not only altered by the tide but also by the
number of months selected as time span. Thus, the water level at Artlenburg,
where tidal movement is almost negligible, gives a more accurate picture of
those spontaneous changes of the Elbe River levels which can be attributed
to climatic events and which can be the only explanation for the changes in
the groundwater level of those by far larger areas of Hamburg where the
Alster River exerts some influence, whereas the Elbe River does not.
Typhus in and the Elbe River Water Level
Hambur~
Typhus Elbe Water Level
Mortality
Hamburg Artlenburg Magdeburg
Year £ermil m m m
1838 15.3
1839 16.1
1840 14.7
1841 17.7 2.1
1842 18.7 1.3
1843 14.1 1.34 1.25 2.3
1844 14.6 1.38 1.87 2.4
1845 14.3 1.39 1.68 1.9
1846 15.5 1.34 1.42 2.0
1847 14.8 1.27 1.41 2.1
1848 11.4 1.17 0.92 1.6
1849 10.7 1.27 1.18 1.8
1850 11.5 1.47 1.62 2.3
1851 11.1 1.43 1.73 2.3
1852 12.1 1.38 1.54 2.0
1853 9.9 1.28 1.82 2.1
1854 10.2 1.47 1.73 2.2
1855 9.6 1.42 2.34 2.3
1856 5.1 1.29 1.44 1.8
1857 16.2 1.10 0.83 1.4
1858 5.2 1.13 0.88 1.5
1859 10.5 1.26 0.97 1.6
1860 5.2 1.34 1.72 2.0
1861 6.6 1.34 1.45 1.8
1862 7.3 1.27 1.28 1.6
1863 6.9 1.34 0.97 1.5
1864 6.3 1.15 0.94 1.4
1865 11.5 1.14 0.85 1.3
1866 9.4 1.27 0.66 1.2
1867 5.0 1.54 1.87 2.1
1868 6.6 1.45 1.45 1.8
1869 6.8 1.33 1.21 1.7
[Continuation on next page}
nico.stehr@zu.de
40 EDUARD BRUCKNER
[Continuation from previous page}
Typhus in Hamburg and the Elbe River Water Level
Typhus Elbe Water Level
Mortality
Hamburg Artlenburg Magdeburg
Year per mil m m m
1870 8.4 Typhus 1.8
1871 5.3 Cases 2.0
1872 6.7 per mil 1.3
1873 5.4 38.1 23 1.3
1874 5.4 23.7 1.0
1875 5.5 32.4 1.3
1876 3.5 15.7 Ground- 1.7
1877 3.0 18.7 water 1.6
1878 3.9 24.7 Hamburg 1.4
1879 2.5 19.4 m 1.8
1880 2.6 15.9 4.90 2.0
1881 3.0 13.9 4.45 2.1
1882 2.7 18.9 3.69 1.9
1883 2.5 13.5 3.48 1.8
1884 2.6 18.3 3.43
1885 4.2 26.8 2.48
1886 7.0 58.1 1.38
1887 8.6 100.7 0.70
The water levels of the Elbe near Magdeburg are not at all influenced by the
tidal range and were mainly included in the study because they are available
up to the year 1883, while the Hamburg and Artlenburg measurements go no
further than 1869. Also included are the groundwater levels near
Hamburg-Eimsbuttel as recorded by C. C. H. Muller since 1880, although
these do not measure up to scrutiny because the surrounding ground was
disturbed.
As it is more difficult to gain an overall picture from tabulations the
results are depicted graphically in the following chart. A variation by I grid
length is equivalent to 1%(l [per thousand] of typhus deaths and 5%0 of typhus
cases, to 0.1 m (one decimeter) of the Elbe water level and 0.5 m of
groundwater variance.
23 For typhus mortality cases the statistical year runs from July 1 to June 30; the annual
average is associated with that year whose first half falls into that statistical period, e.g.,
38.1 in 1873 refers to the period of July 1872 to June 1873, etc.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 41
--
_
;;1;
Mil")
~
--
:;!
r--
;;b
-_
C'I
~
_
~
Mil')
~
..... --
;:Q
r--
~
C'I"'"
~ £ £ :£
M
~ £
or,
f;(; &;
r-
~
C'I"'"
~ &; ~
<"'1
~
.................................................................
II"l r-- 0'1 _ <"'l
_-
1I"lt'-
gg ~
Ii
\
....
-';:1.4' . 1\
r..
~
- - -
..-..............
---
typhus mortality in Hamburg
typhus morbidity in Hamburg
river level at Artlenburg ( A)
_
_
_
'" _ •• -
._.-
river level at Magdeburg ( M)
river level at Hamburg ( H )
--!
-i
,,
--- roundwater in Hamburg i
\ i
II i
11 [;/ IJ , " 1\
f\
,
I
fAr\ ,if
f1 . 1\
A , \ L; IV
.... -
1\ , -,
.'
.
1.1
. .
~
:1'
.
M.
I' 17
"
'.
1'.
.17
I
Figure 1.3
As we look at the table, or rather the chart, the most striking feature is the
tremendous decline of typhus deaths occurring steadily with minor
interruptions over the period from 1842 to 1883. No doubt, this is on the one
hand due to improvements in medical treatments leading to lower mortality
rates and on the other hand due to improvements to the city's sanitation
system leading to an overall decrease of typhus. However, from 1884
onward figures show a rapid and steady increase.
If we compare the three curves representing the water levels, it is obvious
that they follow an almost parallel course. This course is completely parallel
for Magdeburg (M) and Artlenburg (A), despite the fact that they are 230 km
apart. The similarity is much less pronounced between Artlenburg and
Hamburg with only 45 km between them. Though in Hamburg-on the
whole-the Elbe levels show the same fluctuations as upstream they are far
less pronounced. If we take a closer look at the direction these variations
follow over several years, they can be categorized as being parallel, opposite
or indeterminable in the following number of cases:
at Magdeburg and Artlenburg-paraUeI21, opposite 3, indeterminable 2.
at Artlenburg and Hamburg-parallel 18, opposite 8, indeterminable O.
- at Magdeburg and Hamburg-parallel 17, opposite 7, indeterminable 2.
The influence of the tidal range is clearly evident and confirms our
assumption that Elbe River levels at Hamburg can be considered in place of
the groundwater levels for those parts of the city only where the groundwater
nico.stehr@zu.de
42 EDUARD BRUCKNER
is directly affected by the river, whereas the variations of the groundwater in
the Alster region being exclusively the result of climatic events can much
more likely be represented by the Elbe fluctuations at Artlenburg.
In comparing the typhus curve with those of the water levels, we discover
that all of the typhoid epidemics of the years 1842, 1852, 1857, 1859, 1865,
1870, and 1878 occurred when water levels were low, while the years
1843-45, 1860, 1867, 1871, 1876-77 with no typhoid epidemics show high
groundwater levels. Exceptions to this correlation are the minor epidemic
outbreaks of 1846 and 1872, as well as those years 1856 and 1858 with no
typhoid epidemics which show very low levels of groundwater.
If the correlation proves correct, a rise in the groundwater level should
correspond with a decrease in the number of typhoid cases and vice versa. If
one compares the typhus curve with the water levels at Artlenburg, 62% of
all cases seem to confirm the correlation and 38% do not; in the case of
Magdeburg this ratio is 55% and 45%, and for Hamburg 54% and 46%.
According to Soyka for Miinchen, Berlin, Frankfurt a. M., Bremen, and
Salzburg this correlation between groundwater levels and the occurrence of
typhus is on the average indicated in 61 % of all cases. The results are much
more favorable if we concentrate on the more significant variations of the
curve only. A fall or rise of the Elbe water level near Magdeburg of 0.2 m or
more from one year to the next has the opposite effect on the number of
typhus cases in Hamburg in 71 % of all cases; and a rise or fall in the typhoid
curve of 2%0 [per mil] or more corresponds to a counter change of the Elbe
River levels near Magdeburg in 66% of all cases.
In Hamburg, therefore, assuming that fluctuations of the Elbe River level
result in corresponding changes in the groundwater, everything seems to
point to a correlation between groundwater level and typhus in a year-to-year
comparison. At the same time, this seems to confirm our earlier findings
concerning the year-long observations from 1880 to 1883. However, what
about those years 1884 to 1887 when the data did not match the general
pattern at all?
Soyka pointed out that in a long-term view the variations in the
groundwater level from one year to another show a remarkably parallel
development over longer stretches of time, The explanation for this may be
seen in the fact that vast expansive areas experience similar climatic
conditions. I have previously claimed that the rainfall over continental areas
shows a certain general pattern of fluctuations. 24 As a consequence the
groundwater fluctuates in the same way. During the drought of early 1860
which occurred in all of Europe, Asia, America, and Australia, in short all
over the globe, groundwater levels in Europe were low. Since then,
precipitation increased and groundwater levels rose; we may speak of a
24 Annalen der Hydrographie, 1888, February Issue.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 43
change in climate. Some countries on the globe experience spectacular
changes in climate conditions, rainy seasons alternate with fairly dry ones.
Already, the present century recorded three series of wet years around 1815,
1850, and 1880 and some years in between had very little rain.
Groundwater levels changed accordingly and, as we can conclude and as
Penck25 has put it first, so did the conditions of typhoid outbreaks. The
following small table may demonstrate this correlation.
Salzburg Miinchen Elbe River Level Hamburg
Rain Ground- Typhus Rain Ground- Typhus Artlen- Magde- Typhus
water water burg burg
mm m 0/00 mm m 0/00 m m 0/00
1841/45 2.00 15.9
1846/50 1.31 1.96 12.8
1851/55 1.83 2.18 10.6
1856/60 781 0.09* 23.6 1.17 1.66 8.4
1861/65 1058* 2.97* 755* 0.27 18.7 1.10* 1.52* 7.7
1866/70 1207 3.03 9.3 790 0.42 12.0 1.0 1.76 7.2
1871175 1283 2.99 14.9 766 0.27 15.4 1.38 5.6
1876/80 1431 3.13 5.9 874 0.75 7.2 1.70 3.1
1881/85 1285 3.04 2.6 919 0.35 1.7 1.93 3.0
It is highly regrettable that concerning the occurrence of typhUS the
morbidity rate is not available to us over a more extended number of years,
but that we have to rely on the mortality rate instead. These numbers are
misleading whenever long time-spans are involved since they are lowered
considerably through progress in medical treatment. Only the actual number
of cases can serve as indicator for the frequency in which illnesses occur.
That is why in Hamburg, during those years of 1856 to 1865 when river and
groundwater levels were low, the typhus death rate did not rise nevertheless.
It seems likely that the typhus infection was even higher than during
previous lustra: the fact that the general decline of the mortality rate slowed
down significantly during those very lustra makes me believe that in fact the
morbidity rate was higher. From one lustrum to the next mortality rates
declined as follows:
1846/50 51/55 56/60 61/65 66/70 71/75 76/80 81/85 86/87
3.1 2.2 2.2 0.7* 0.5* 1.6 2.5 0.1 4.7
In conclusion, there may have been an overall increase in typhus cases in
the Sixties indicated by a stagnation in the decline of deaths, while the more
likely time-span with high water levels around 1850 and at the end of the
25 Miinchener Allgemeine Zeitung, 1887.
nico.stehr@zu.de
44 EDUARD BRUCKNER
Seventies shows a rapid decline of mortality. Similar conditions were found
in Munchen.
As of 1881 precipitation in Hamburg as well as in all the rest of central
Europe declines. Hamburg's rainfall measured in millimeters was:
1876/80 1881 1882 1883 1884 1885 1886 1887
867 695 673 586 758 620 598 466
This resulted in a general drop of the groundwater levels, as for instance
Soyka reported for Bremen, Berlin, Frankfurt a. M., Munchen, and
Salzburg. 26 Hamburg, too, experienced an enormous drop by a full 4.2
meters as C. C. H. Muller's data verify. However, we also know that the
groundwater level of Eimsbuttel sank considerably due to the construction of
the Isebeck canal in 1883 and 1884, which might explain most of the
unprecedented reduction by 4.2 m in the course of seven years. Still, in part
this reduction can most definitely be attributed to the long periods of little
precipitation over these years, which, consequently, should be the case for
all of Hamburg. It appears only logical to make a connection between this
drop in groundwater level and the major epidemic in Hamburg during the
years 1884 to 87. If this would be the case, one would expect epidemic
outbreaks of typhus in other cities as well where the groundwater level sank
significantly. That, however, is not the case, except for Salzburg and
Frankfurt a. M. which show a very small increase of typhus cases during a
period of dropping groundwater levels in the years from 1883 to 1895.
Berlin and Munchen had no such increases, neither did Hamburg's
neighbouring city of Bremen in particular, although its water level dropped
by 49 cm during 1881-84. 27 Soyka emphasizes this lack of consistency in the
correlation between groundwater and typhus during the Eighties and sees its
cause in the steady progress of improving the underground sanitation system,
which ultimately will eliminate typhus altogether. 28 Could it be assumed that
such progress did not take place in Hamburg so that the drop of the
groundwater level could cause an epidemic of proportions which by the
sheer number of cases might be called unparalleled in the city annals? This is
quite unlikely, given that Hamburg is one of the first large cities in Germany
that with its system of sluices created a magnificent drainage system.
Therefore, it seems probable that the cause of Hamburg'S last typhus
epidemic is to be found chiefly in some local conditions peculiar to
26 Schwankungen des Grundwassers [Fluctuations o/Groundwater], pp. 83.
27 Not by 2.96 meters as Soyka has stated, in Schwankungen des Grundwassers, p. 78; he
neglected to divide the sum of the annual averages of the different wells in Bremen by the
total number of wells in order to obtain the overall average. This mistake has been
corrected in Figure 2.
28 Archive fUr Hygiene, Vol. VI, p. 282 a. 285.
nico.stehr@zu.de
GROUNDWATER AND TYPHUS 45
Hamburg and is not linked to groundwater levels in general. This view is
confirmed by the fact that, as previously mentioned, the seasonal outbreak of
typhus does not conform at all with seasonal fluctuations of groundwater.
These past years of epidemics, therefore, show an abnormal relation between
groundwater and typhus.
But what is the possible local cause of the epidemic? "Could it be
chance", as I stated less than a year ago,29 "that the beginning of the
epidemic coincides with the beginning of the harbour construction work for
the purpose of joining the customs-union? If this massive excavation of soil
and dirt was indeed the cause, future will tell. Let us hope so, because in that
case we may be about to see the end of the epidemic as harbour construction
is about to be completed." Some facts actually support this assumption. The
soil movements at the construction sites were completed during the first six
months of 1888 and in October of that year the new docks were opened. At
the same time, the typhus epidemic was considered retrogressive since the
spring of 1888. Of course, the groundwater level rose slightly as well in the
wet year of 1888, which weakens this argument somewhat.
Let us review the results of our efforts. In Hamburg, too, the outbreak of
typhus is closely related to the fluctuations of groundwater levels. This
correlation is quite evident when we compare the annual changes of both
factors as observed during 1880-83. Even the changes from one year to the
next confirm this relation, as indicated by the fluctuations of the Elbe river
levels which were used from 1841 on because proper measurements of the
groundwater levels were unavailable.
However, in the epidemic years of 1884-87 this correlation is strikingly
absent. The annual pattern is different: what used to be a summer disease has
turned into a winter disease. In addition, the outbreak has grown to shocking
unprecedented proportions. It cannot be connected to the drop of
groundwater level as this was the case in all of central Europe without
causing a widespread epidemic of remotely similar proportions. Rather this
epidemic seems to originate from massive soil movements in connection
with the new harbour construction in Hamburg because it runs its course
over exactly the same length of time that soil movements take place and
vanishes after they have stopped. During construction, vast masses of moist
and drying soil containing numerous bacteria were exposed to air enabling
the spread of these carriers of disease to the human body. Whether now, after
digging is completed and the epidemic has passed, the occurrence of typhus
will again, annually as well as from year to year, show the same pattern of
variation as the groundwater level, future research will tell.
29 Hamburgischer Korrespondent 1888, No. 85,25 March, Morning Issue, p. 10.
nico.stehr@zu.de
Chapter 2
Fluctuations of Water Levels in the Caspian Sea, the
Black Sea, and the Baltic Sea Relative to Weather*30
The contemporary opinion, more or less commonly shared by hydrographers
and oceanographers, is that the fluctuations of water levels, as observed by
water-mark stations at seas which are fairly isolated from open ocean waters,
are attributed to the influence of wind. And indeed, if we keep in mind the
powerful tidal waves that, under the impact of northerly and northeasterly
winds, haunted the lower German coasts repeatedly, this view would seem
plausible. As easy as it is to verify and substantiate the direct effect of wind
on water levels and their irregular changes from day to day, the more
complicated it is, however, to interpret the slow change of the average water
levels to be the direct result of the wind. For it was noticed that apart from
the short-term irregular variations that occurred differently at different
coastal points of the same oceanic waters, long-term general fluctuations
take place as well. The Baltic Sea, for example, where average water levels
of different months were compared, shows a distinct annual periodicity
whose course appears to be more or less parallel at all measuring stations
despite their distance. Fluctuations of this kind could no longer be consi-
dered a consequence of local winds since they indicated a considerable
change in the water volume of the Baltic Sea. General wind conditions and
atmospheric wind pressure-which at times pushed the waters of the Baltic
Sea towards the South and the West, those of the North Sea towards the
Die Schwankungen des Wasserstandes im Kaspischen Meer, dem Schwarzen Meer und der
Ostsee in ihrer Beziehung zur Witterung. Annalen der Hydrographie und Maritimen
Meteorologie, Vol. II, 1888.
30 Lecture held at the General Assembly of the German Meteorological Society in Karlsruhe,
Easter 1887. The research and its results explained below is part of a larger essay by the
author, with the title Klimaschwankungen seit 1700 [Climatic variations since 1700; three
chapters of this extensive work are presented in this collection in Chapter 4] which will be
published by the end of 1888 in the III. volume of "Geographische Abhandlungen", Wien,
Holzel. Some important addenda to this lecture are the result of additional research by the
author and have previously appeared in the commemorative papers.
47
nico.stehr@zu.de
48 EDUARD BRUCKNER
West, at other times in tum let the North Sea rise eastward and the Baltic Sea
surge towards the North and East---<:ould not be the only consideration here.
Baensch31 and H.A. Meye~2 explained the fluctuations of water levels in as
far as they are the result of a volume change this way: northerly and easterly
winds cause an increase of flowing water masses into the North Sea, and a
decline of water mass in the Baltic Sea; westerly winds on the other hand
cause a reduction in the outflow and a rise of the Baltic Sea level. Tempera-
ture, atmospheric pressure and precipitation should have no immediate effect
on average water levels.
But in contrast to this opinion, Seibt showed in 1885 that the changes of
the water-level in the course of one year as well as from year to year do not
at all correspond nicely with prevailing wind conditions, as one was inclined
to assume: 33 thus, wind could no longer be held responsible as the sole cause
of those fluctuations. Seibt thinks the annual seasonal timing could be traced
back to a yearly recurring tidal wave. In the summer, movement of water
masses of the global oceans occurs from the south to the north, in winter the
movement is reversed. Naturally, such a shift would alter water levels inclu-
ding those of seas or lakes interconnected with the oceans. According to
Seibt and v. Baeyer4 this annual tide cycle, which follows the cycle of the
sun, can only be observed in the adjoining oceanic waters, while open
oceans with a twice daily tide cycle entirely defuse variations. Maydell de-
notes a similar view in 1884 as being insupportable and useless in defining
the annual periods of water level changes in seas with no outlet: he was at a
loss to explain the annual variation in the Black Sea. 35 This is noteworthy,
because at the same time he was the first to prove that the year-to-year
change of the Pontus' [old-fashioned for Black Sea] medium water levels
has to be regarded as the direct result of the change in the annual precipita-
tion and, consequently, of the related water volume of the rivers. I may be
permitted to demonstrate the overwhelming importance of this last factor,
which even Maydell underestimated by far, namely the rivers' capacity, in
the water levels of three European seas: the Caspian Sea with no connection
31 Studien aus dem Gebiet der Ostsee [Studies for the Baltic Sea Region], 1872.
32 Untersuchungen iiber die physikalischen Verhiiltnisse des Westlichen Theiles der Ostsee
[Research ofphysical conditions ofthe western region of the Baltic Sea].
33 Das Mittelwasser der Ostsee bei Travemiinde [The mean water level of the Baltic Sea at
Travemiinde], Publikation des Koniglich Preuftischen Geodiitischen Instituts, Berlin 1885.
34 see Meteorologische Zeitschrift J885. November Issue.
35 Morskoj Sbornik 1884. Vol. 11. November. In an essay of Currents and Water Exchange
between the Black Sea and the Mediterranean Sea, published in 1886 in the
Meteorologische Zeitschrift, p. 532. Not quite correctly it is pointed out that v. Maydell
links the seasonal [depending on the months] changing water levels of rivers to the annual
fluctuations of the sea's level: this was my own view mentioned in an essay of the
magazine "Naturforscher" on February 27th, 1886--0pposing v. Maydell's opinion.
nico.stehr@zu.de
FLUCTUATIONS OF WATER LEVELS 49
to the ocean; the Black Sea only connected by the narrow and shallow
Bosporus with the Mediterranean Sea and thereby with the ocean; and the
Baltic Sea with a more open connection.
2.1 The Annual Water Level Cycle
In order to substantiate the annual cycle of water levels in the Caspian Sea
we have two series of observations which were carried out at Baku (over 25
years) and at Aschur-Ade in the bay of Astrabad (17 years).36 I determined
the medium values of the two quite congruent series and obtained the
following numbers. The + (plus) sign indicates a water level above the
annual average, the - (minus) sign a level below the average. We encounter
the lowest water level in March and the highest in July and August. The
amplitude of variation is more than 0.3 meters and the water volume of the
Caspian Sea, determined on the basis of this amplitude and the Sea's entire
area, changes by 165 cbkm (cubic kilometres).
Jan. Feb. Mar. AEr. Ma~. Jun.
Caspian Sea cm -14 -15 -16* -8 0 13
Volga cm -50 -60 -70* -40 120 220
{Astrachant
Jul. Au~. SeE' Oct. Nov. Dec.
Caspian Sea cm 21 22 15 2 -9 -11
Volga (Astrachan) cm 100 -20 -40 -30 -50 -60
That this large variation recurring year after year with near absolute
regularity is caused by the annual periodicity of the water supply from
several rivers, predominantly the Volga, is rather obvious. Indeed, a consen-
sus between the Caspian Sea and the monthly measured water levels of the
Volga near Astrachan does exist: at both places winter is the season of low
levels and summer of high levels. Noteworthy is a very distinct delay in the
peak level, which the Volga reaches in June after the snow melts in Northern
Russia, and the Caspian Sea 1 112 months later in July/August. Yet it is this
delay that confirms the causal relationship between the two phenomena:
because the delay of the maximum water level in the basin of the Caspian
Sea is a logical consequence if its level changes are the result of changing
Volga levels. The water level of the Sea keeps on rising even after the actual
floodwaters of the Volga river are fully dispersed, until in late summer the
increasing evaporation off the water's surface reduces the level: water
accumulates in the Sea.
36 Published by Filipof, Variations of Water Levels in the Caspian Sea, Morskoj Sbomik,
1880 No. 7 and 8 (in Russian).
37 According to Woeikof, Klimate der Erde [Climates ofEarth] (Jena 1887), Table XXII.
nico.stehr@zu.de
50 EDUARD BRUCKNER
The correlation between the annual fluctuation of the water level in ba-
sins connected to the ocean and the annual of their tributaries is not at all as
clear. Using the same numbers that v. Maydell was unable to explain,38 r can
prove qualitatively and quantitatively that the annual water level of the Black
Sea is also dependent on the rivers' discharge. 39 The following tabulation of
figures from the Black Sea, the Donau and the Don will outline this correla-
tion.
Jan. Feb. Mar. AEr. Ma~ Jun.
Don at Kliiatsch 40 -0.7 -0.4 0.6 34 37 0.4
40
Dnjepr at Krementschug -0.4 -0.4 0.4 23 26 0.8
40
Donau at Orgsova -0.5* -0.3 0.4 1.0 1.0 0.6
Pontus at River Delta -0.061 -0.056 -0.013 0.075 0.130 0.107
Pontus at coasts without rivers -0.053 -0.058* -0.036 0.048 0.109 0.130
Jui. Aug. SeE· Oct. Nov. Dec.
Don at Klilatsch40 -0.4 -2.0 -2.4* -1.7 -0.7 -0.7
40
Dnjepr at Krementschug -0.7 -1.1 -1.3* -1.2 -0.9 -0.7
Donau at Orgsova40 0.0 -0.5 -0.8 -1.0* -0.5 -0.2
Pontus at River Delta -0.048 -0.010 -0.058 -0.058 -0.051 -0.051
Pontus at coasts without rivers 0.094 0.020 0.05 -0.084* -0.079 -0.038
Again, the time delay caused by the accumulated amounts of water col-
lected in the Pontus basin is evident: the highest river-water level occurs in
April and May and the highest level of the Black Sea in May/June.
Another secondary, less dominant, but not all that insignificant factor in
reaching the summer's maximum level is the thermal expansion of the Sea's
water volume by 44 cbkms t as a result of the temperature rise between
February and May.
The conditions in the Baltic Sea are considerably more complicated due
to its freer water exchange with the ocean. Despite the sinuous shape of this
basin, the variations throughout the year run parallel at various points along
the coast. Generally, levels are at their highest in the summer: after that the
sea-level falls and reaches its lowest mark in the spring. This periodicity is
not quite as constant at these different measuring stations if we compare
monthly averages instead of seasonal ones; it is however constant enough for
same time spans, so that one station may serve as representing the entire Sea.
r have selected Swinemunde for this purpose, since for this station there are
carefully researched and absolutely reliable data at hand prepared by Seibt.41
38 Op. cit.
39 Compare: Meteorologische Zeitschrift, 1886, July Issue, p. 297.
40 By Woeikof, Klimate der Erde [Climates ofEarth], Jena 1887, Figure XXII.
t [cubic kilometres]
41 Miftelwasser der Ostsee bei Swinemunde [Medium Water Level of the Baltic Sea at
Swinemunde], Berlin 1881.
nico.stehr@zu.de
FLUCTUATIONS OF WATER LEVELS 51
I am not asserting a general medium value but three decadal means for the
decades 1855-64, 1865-74, and 1875-84. It is, as emphasized by Seibt,
quite remarkable that these three decades show three very different annual
curves: the mean of 1855-64 has no November maximum at all, which in the
1865-74 decade is more pronounced than the August maximum and which
to a lesser extent also occurs in 1875-84.
The significant change from one decade to the next is an indication that
the annual periodicity of the water level in the Baltic Sea is influenced by a
number of factors, each with a different emphasis at different times.
Baltic Sea at Swinemiinde {mm}
Jan. Feb. AEr. Ma~ Jun. Jul. Au~.
Mar. SeE· Oct. Nov. Dec.
1855-64 13 2 -38-25 -48 -53 86 111 57 3 -45 -74*
1865-74 -38 -56 -67
-72* -29 25 54 54 34 -30 87 35
1875-84 -16 -28 -13
-86* -71 32 62 69 45 -2 17 -12
Oder River at Kiistrin42 (cm)
1807-35
_ .. _
15
--
39
---
57 50
------
10 -22 -23 -35 -32 -40*
- - - - - - - -
-28 -10
G10mmen1l2
1862-76 -99 -114 -123* -78 78 226 119 54 43 12 -65 -94
A comparison between the annual fluctuation of the Baltic Sea level and
that of the water level of the major German rivers, in the table above
represented by the Oder River, indicates that the rivers' run-off can only play
a very subordinate role: the Baltic Sea's minimum level coincides with the
rivers' maximum levels. Scandinavian rivers, represented here by the
Glommen river, show a quite different pattern. Their summer maximum may
have some impact on the maximum level in the Baltic Sea, which occurs at
almost the same time, yet is by no means solely the result of it. The annual
amount of precipitation certainly plays an important role increasing the
Baltic Sea's water volume to its maximum level during the summer.
Another accountable factor for the high water level in the summer is the
thermo-dynamic increase of the Baltic Sea's water volume due to its warmer
summer temperatures. This factor alone causes the volume to increase by
30 cubic kilometres or the level to rise by 8 cm. We must remember that
evaporation increases during the summer as well, so that water lost through
evaporation counteracts the volume expansion by a certain extent.
However, all these factors do not explain the Baltic Sea's minimum level
in springtime, which could well be related to the easterly and northeasterly
winds prevailing at the time of its occurrence. The high water level in the
summer may also partly be attributed to the steady westerly winds at that
42 According to Wex: Ober die Wasserabnahme in den Quellen und Stromen [About water
decrease of springs, streams, and rivers], Zeitschrift des osterreichischen Ingenieur- und
Architekten-Vereins, 1879.
nico.stehr@zu.de
52 EDUARD BRUCKNER
time of the year. Northeasterly winds force water out of the Baltic Sea while
westerly winds have the reverse effect.
We are therefore faced with a multitude of aspects all of which, without
any doubt, influence the annual periodicity of the Baltic Sea's level changes;
however, at present, we are in no position to weigh their quantitative impact.
Results tum out differently when comparing the variations of water levels
from one year to the next.
2.2 SECULAR WATER LEVEL VARIATIONS
Rarely does a sea level at the end of a year return to the same level it used to
have at the beginning of that year: rarely are water input and water output in
equilibrium over the course of one year. Ordinarily, one of the two factors
outweighs the other in the annual balance, resulting in a rising or falling
water level: the average water level varies from year to year. It is now our
intention to observe this change in the three seas selected.
Fluctuations in the medium water levels from one year to the next are
quite substantial and this is quite natural in a basin with no outlet such as the
Caspian Sea where water is supplied by precipitation and its tributaries and
discharged through evaporation. These fluctuations are all the more pro-
nounced since the weather tends to have the opposite effect on water supply
and discharge: rainy years adding to the water volume tend to be cool with
little evaporation and little water loss.
It becomes evident that these fluctuations of the Caspian Sea do not level
off over a period of a few years, but rather that certain major fluctuations of
the water level take place over longer periods of time. Let us look at the lus-
tra' averages of water level observations at Baku and Aschur-Ade published
by Filipof;43 after being adjusted in relation to a fixed point zero as a
common denominator, these figures show that from 1851 to 1865, water
levels of the Caspian Sea were low, and began to rise steadily from
1866-overall by approx. 3/4 of a meter. This rise is equally prominent at
both stations and must be considered an absolutely certain fact. The
conformity between both stations contradicts beforehand the assumption that
this fluctuation is caused by movements of the ground formation, as one
might be inclined to assume of the Baku area rich in naphta deposits. The
mean of both observation series is shown below as well as in Figure 2.
Caspian Sea
1851-55 185~0 1861-65 1866-70 1874--75 1876-78
em -21 -26 -19 -19 +16 +56
, [period of five years]
43 op. cit.
nico.stehr@zu.de
FLUCTUATIONS OF WATER LEVELS 53
These observations series are not long enough to show similar long-term
movements of the Pontus level. The data of the years 1874-1882 reveal
however, that the medium water level changes from year to year varying up
to 36 cm during this time span. When we take a closer look at the main
characteristics of this movement we will again recognize a rise of the water
level towards 1880 as was the case with the Caspian Sea.
The lustra means of the Baltic Sea as well reveal long-term water level
variations, the amplitude of which is however measured in decimeters and
not in centimeters as is the case with the Pontus and the Caspian Sea. These
are too pronounced and too universal to be the result of observational errors
or changes of the measuring posts.
In order to trace these fluctuations, I determined lustra averages for 10
German water level stations which are, ranging from East to West, Memel,
Neufahrwasser near Danzig, Stolpmunde, Wiek near Greifswald, Stralsund,
Warnemunde, Wismar, and Travemunde, shown non-adjusted in the
following table and depicted graphically in Figure 1.
:1
Meme
Stolpmlinde
Neufahrwasse
Colbergermlinde
Swinemlinde
Wiek
Stralsund Warnemlinde
Wismar
Travemlinde
1846 50 51 55 56 60 61 65 66 70 71 75 76 80
Figure 2.1
It is certainly a drawback that, with the exception of these German
stations, no long-term series of water level observations are available at
present. However, the number of existing stations in various locations is
sufficient and observations at the coastal regions of Sweden and Finland not
included here, s~em less comparable since those coastlines are supposedly
nico.stehr@zu.de
54 EDUARD BRUCKNER
undergoing a secular change in elevation which might distort the
independent secular movement of the water level. 44
A glance at the table and more so at the curves in Figure 1 tells us that
the movements of the water level are similar at all stations: a decline starting
in 1850 or 1855, reaching a minimum from 1860 to 1865, another increase
beginning in 1866, then a sudden jump forward at first followed by a slight
decline in 1871-75 and a gradual rise. It is quite remarkable that in the Baltic
Sea we again see this rise of the sea level starting in the middle of the
Sixties, which we encountered at the Caspian Sea.
Unfortunately, there are not enough precise observation series available
to trace secular water level variations far back in time. Only the Swinemiinde
station provides a series of observations starting at the end of the Twenties,
which tell us that the general maximum around mid-century (1845-55) was
preceded by a period of low Baltic Sea levels (cf. Figure 2).
Baltic Sea (mm)
'""'"...
~
<I)
:g
'" :;:1
~ ~ :g
00
I
til
til
1]
:§
'Co
~
r--
"'"...«!
:;:1 :;:1
'""'"
J
:;:1
E E
] E 'Co
oS <I) <I)
~ ~
~
.E- ::9 .5 Number of
as S o
Lustrum ~ z VJ u ~
VJ ~ VJ ~ +
1826-30 -7
1831-35 -5
1836-40 -5
1841-45 -6
1846-50 +19 -4 +25 +9 -0 +26 +12 +47 +24 9
1851-55 +11 +23 -5 +7 +19 +23 +17 +10 +33 9
1856-60 -20 +2 -30 -32 -14 -3 -13 -23 -23 -20 9
1861-65 -17 -23 -23 -28 -33 -24 -32 -21 -25 -29 10
1866-70 +28 +16 +46 +24 +41 +0 +43 +19 +19 +10 10
1871-75 -19 -15 -II +15 -12 -18 -26 -12 -12 -19 9
1876-80 +47 +38 +11
As far as the Caspian Sea is concerned matters are much better. Although
level observations do not even date back further than the middle of the
44 Later on I was able to verify similar long-term variations of water levels in these coastal
regions as well.
45 According to Levels and Height Adjustments, etc., made by the Trigonometrische
Abtheilung der Landesaufnahme. Vol. III. § II, p. 139. Berlin 1875.
46 According to Seibt, Miltelwasser der Ostsee bei Swinemiinde [Medium Water Level of the
Baltic Sea at Swinemiinde]. Berlin 1881.
47 As in footnote 15 but supplemented by reports of the Kiel Commission.
48 According to Seibt, Miltelwasser der Ostsee bei Travemiinde [Medium Water Level of the
Baltic Sea at TravemiindeJ Berlin 1885.
nico.stehr@zu.de
FLUCTUATIONS OF WATER LEVELS 55
current century, the water level variations, occurring here on a far grander
scale, have caused radical changes to the coastlines: there are, for instance,
islands which at times vanish and reappear above sea level, at times
reconnect with the mainland turning into peninsulas, and then again reappear
as proper islands. Consequently, these variations have attracted the inhabi-
tants' attention from very early on. In fact we have numerous reports which
include valuable data concerning the changes of the sea level, critically as-
sessed by Lenz and Sokolof. I do not agree with Filipof,49 who denotes the
results of both researchers as being not sufficiently verified. They corre-
spond very well with other publications so that I have to consider them quite
reliable and sound. 50 The sea level changed as follows:
CasEian Sea
Movement Hi~hLevel Low Level
Until 1744 Rise about 1745
Until 1766 Fall about 1765
Until 1809/14 Rise about 1815
Until 1842 Fall about 1845
Since 1847 Rise about 1850
Followed By Fall about 1860
Since 1866 Rise (about 1880)
At this point let us summarize what we have discussed so far. Both, the
Caspian Sea and the Baltic Sea show level oscillations; long-term periods of
rising levels alternate with falling ones. As far as we can tell from observa-
tions available, the rise and fall occurs at both seas at the same time and in a
parallel pattern, whereby, though, the amplitude of the Caspian Sea's oscilla-
tions is far more pronounced than the one of the Baltic Sea level.
A comparison between these variations and those of glaciers seems obvi-
ous since seas with no outlet, such as the Caspian Sea, and glaciers are in
size and extent regulated by the same factors: precipitation, which in both
cases supplies water, and warm temperatures, which in the one case evapo-
rate the water and in the other melt the ice. As much as we know about
glaciers, the variations of the two phenomena do indeed show a parallel
pattern. The period of advancement by the glaciers of the Alps at the begin-
ning of the century corresponds with a period of rising and high water levels
of the Caspian Sea. The second period of glacial expansion at the end of the
Forties is paralleled by rising Caspian Sea levels, and the most recent
increase in the Caspian Sea level starting in 1866 finds its counterpart in the
49 Op. cit.
50 A more detailed argument is mentioned in another place. Additional references originated
only in minor changes of the table. In tum these variations were traced back to 1685. 1685
to 1715 decline; 1715 to 1744 rise, also high level in 1685, low level in 1715.
nico.stehr@zu.de
56 EDUARD BRUCKNER
glacial advancement developing since the late Seventies. Glacial change
seems to lag behind the change in the Caspian Sea level by a number of
years. What Swarowsky demonstrated for the glaciers of the Eastern Alps
and the Neusiedler See51 is repeated, if on a much grander scale at the
Caspian Sea. Both phenomena, glacial variations and variations of seas with
no outlet, go back to the same cause, to secular weather changes, to climatic
changes.
Weather affects the water level of lakes and seas with no outlet by go-
verning the input and output of water. Since long-term observations of river
levels, which would show any variation in the river volume directly, are
unavailable for the area of the Caspian Sea, I shall have to revert to precipi-
tation measurements. 52 St. Petersburg in the Northwest of Russia, Lugan in
the South, Bogoslowsk in the Ural and Tiflis in the Caucasus shall represent
the different parts of the Caspian Sea's catchment area. Although some of
these measuring points seem far away from the Caspian Sea or even beyond
its boundaries, we are nevertheless entitled to include them in our inquiry
since, by experience, secular weather changes tend to affect large territories.
At all stations we see a decline in rainfall starting around the late Forties
and early Fifties and continuing into the Sixties when it starts to increase
considerably. These changes in precipitation within the catchment area of the
Caspian Sea, become more apparent if we express the lustrum averages of
each station in percentages of the mean value of the years 1841-1880 on the
basis of which we can then determine lustra averages for the entire area. If
we compare these numbers in the table and the diagram of Figure 2 with
those of the level variation of the Caspian Sea, there can be no doubt that the
latter is caused by variations in precipitation.
It seems of special interest to find the causes for secular level changes of
the Baltic Sea because in this case the factors that might have an effect are so
numerous. Our findings about the Caspian Sea lead us to an instant compari-
son between the variations of the Baltic Sea and those of its surrounding
tributaries. Long-term water level variations can also be observed in Central
European rivers, which allows us to draw direct conclusions about volume
fluctuations. It is astonishing that these variations have received little recog-
nition so far, although Hagen53 and above all Fritz54 have called attention to
it. The following tabfe [following page] shows the lustra averages, based on
51 XII Bericht des Vereins der Geographen an der Universitiit Wien, 1886.
52 Observations of the Volga river level near Astrachan subsequently made available to me,
fully confirm these findings.
53 Wasserstiinde preussischer Strome [Water Levels ofPruss ian Rivers]. Sitzungsberichte der
Berliner Akademie der Wissenschaften, 1880.
54 Petermann's Geographische Mittheilungen, 1880, p. 245.
nico.stehr@zu.de
FLUCTUA TIONS OF WATER LEVELS 57
data collected by Fritz,s5 of the medium water levels of the Memel near Tilsit,
the Weichsel near Kurzbracke, the Oder near Kiistrin and Neuglietzen, also
from the greater vicinity of the Baltic Sea those of the Elbe near Magdeburg,
the Weser near Bremen, the Rhine near Diisseldorf and Emmerich, and the
Seine near Paris. The [rightmost column] refers to the Donau at Orsova.
These numbers are charted in Figure 2 so that a change by one graduation
mark equals a change of the water level by 0.2 meters. The two series for the
rivers Oder and Rhine respectively were integrated into one curve each.
Rainfall (mm) 1841- 1846- 1851- 1856- 1861- 1866- 1871- 1876-
45 50 55 60 65 70 75 80
St. Petersburg 482 436 369 350* 443 603 517 518
Lugan 385 287 352 311 280* 379 384 467
Tiflis 522 454 452* 465 530 448 520
Bogoslowsk 387 452 412 384 320 282* 483 466
Mean 103 89 94 88* 88 105 108 117
Caspian Sea -21 -26* -19 19 16 56
Water Level
(em)
From the table and even more so from the graphic chart in Figure 2 it is
evident that the changes of the water levels of those rivers show in all of
Central Europe an almost parallel pattern from lustrum to lustrum. All of
these rivers reach their highest levels around 1801-10 which then continue
to fall more or less uninterruptedly. Grouped around the year 1830 are the
lowest lustra averages. By 1850, or to some extent earlier, we encounter a
second maximum, which occurs in the same universal fashion as the first. A
very steep decline follows and, almost without exception, a minimum is
reached during the years of 1856-65. Since 1866, rivers are once again on
the rise, a period which does not seem to have come to a close yet in 1880.
Generally, in Central Europe this last variation is more pronounced than the
one from the first half of the century.
A comparison between these variations of river levels and the variations
of the Baltic Sea level, shown in the Swinemiinde measurements of Figure 2,
reveals a striking parallel pattern of the two and the surprising finding that
the water level of the Baltic Sea rises during years when the water supply
from rivers is high, and falls when the situation is reversed. This parallel
pattern is so perfect that chance can be excluded entirely. Even the intensity
of the fluctuations is alike: a less pronounced variation in the first half of the
century (Swinemiinde), and a more prominent one during the Sixties and
Seventies.
55 The link between sun spots and magnetic and meteorological phenomena on Earth. Harlem
1878,p.135.
nico.stehr@zu.de
58 EDUARD BRUCKNER
It must be determined, whether both events have one and the same cause
or whether one is the cause of the other. A common source might perhaps be
the wind. To a large extent all precipitation is dependent on wind conditions
and the rainfall determines the amount of water the rivers carry; stronger air
currents carrying rain must therefore increase the rivers' volume and raise
their levels. Also no doubt, wind and storm conditions intensify the drift
between Baltic Sea and North Sea either reducing or increasing it. But it is
these rain-carrying storms, the westerly winds in particular, which are
holding Baltic Sea waters back, while the dry winds from the continental
masses support the Sea's outflow through the belts and the Sound. Thus, it
would be possible and conceivable to hold wind conditions indirectly
accountable for the fluctuations of river levels and directly accountable for
fluctuations of the Baltic Sea level. However, a comparison between the
figures and charts of individual river and sea measuring stations advises
otherwise. The water level of a particular part of the Baltic Sea is more
closely related to its neighbouring large waterways than to its more distant
parts. This is shown in the following table. With regard to the first maximum
and minimum, the Neufahrwasser data reveal a delay in comparison with
those of its nearest measuring stations Memel in the East and Stolpmiinde in
the West, yet correspond in this regard to those measurements taken at the
Weichsel station Kurzbracke. 56
1846-50 1851-55 1856--60 1861--65 1866-70 1871-75
Memel mm +19 -11 -20* -17 +28 -19
Stolpmiinde mm +25 -5 -30* -23 +46 -11
Neufahrwasser mm -4 +23 +2 -23* +16 -15
Weichsel mm 187 231 140 94* 152 168
The figures demonstrate the direct link between river volume and the
Baltic Sea's water level. They do not reflect the impact of the wind condi-
tions which are difficult to measure since they are constantly changing.
Based on the correlation between the amount of rainfall and the level of
the Pontus in the years 1874 to 1882 57 it is safe to say that the secular
variations of the water levels of the Caspian Sea, the Baltic Sea and the
Black Sea are the result of variations in the water volume of the rivers
feeding into these seas. This latter fact points to secular variations of the pre-
cipitation-to secular climate variations. Similar variations in precipitation
and temperature were identified for the Alps by Lang in connection with his
most valuable inquiry into glacial changes. 58 It cannot be a mere coincidence
56 This fact is even more evident when we compare the variation of the Baltic Sea level along
the Swedish and German coast.
57 v. Maydell and other authors.
58 Zeitschrift fUr Meteorologie, 1885, p. 113 (433).
nico.stehr@zu.de
Secular variations of water levels of central European rivers (m).
Memel Weichsel Oder Elbe Weser Rhein Seine Donau
Tilsit Kurzbracke Kustrin Neu-Glietzen Magdeburg Bremen Dusseldorf Emmerich Paris Orsova
1766170 2.91 1.24 '"Ti
1771175 2.95 3.52 1.40 r
1776/80 2.63 3.24 1.20
c:::
(1
1781185 1.55 2.72 3.09 1.12 >-i
1786/90 1.61 2.47 3.18 1.32 ~
1791195 1.17* 2.15* 3.08 1.22 >-i
1796/1800 1.33 2.23 2.74* 1.09* 0
-
Z
1801105 1.67 2.54 2.76 3.10 1.32 en
1806/10 1.33 2.13 2.98 3.02 1.37 0
'Tl
1811115 2.32 1.12 1.78 1.88 2.48 2.62 1.11
1816/20 2.46 1.28 1.88 1.97 1.33 2.93 2.90 1.46
1821125 2.11 1.00* 1.66* 2.01 1.14* 2.74 2.82 1.14
~
>-i
tr:l
1826/30 2.20 1.37 2.06 2.25 1.56 2.65 2.69 1.10 :;0
nico.stehr@zu.de
1831/35 1.67* 1.11 1.71 1.83* 1.22 2.48* 2.41 * 1.04* r
1836/40 2.14 2.23 2.02 1.38 2.88 1.49 tTl
1841145 2.01 2.04 2.05 1.47 2.85 1.29 3.04
<:
tTl
1846/50 2.37 1.87 2.39 1.99 1.11 2.74 1.26 2.97 ren
1851155 2.67 2.31 2.69 2.21 1.44 2.93 1.43 3.15
1856/60 1.89 1.40 1.89 1.71 0.69* 2.29 0.93 2.54
1861165 1.77* 0.94* 1.76* 1.52 0.88 2.20* 2.19*
1866170 2.06 1.52 2.06 1.76 1.02 2.67 2.65
1871175 2.00 1.68 2.12 1.38* 0.83 2.48
1876/80 2.35 1.65 2.35 1.70 2.66
VI
\0
60 EDUARD BRUCKNER
that the weather changes we discovered through water level observations at
seas and rivers occurred at exactly the same time as those observed by Lang
in the Alps.
Let us take a closer look at the entire area that, as was established earlier,
is affected by the climatic variation (see Fig. 2): it includes the catchment
areas of the Caspian Sea, the Black Sea,59 the Baltic Sea, the North Sea, also
the Seine-basin and the Alps with the Po delta, i.e., nearly all of Europe.
Thus, we have to conclude that almost all of Europe experiences climatic
variations at the same time and in the same manner.
The immediate question arises where the territorial boundaries of these
climatic variations may lie, whether they are limited to Europe or extend
completely or partly to other parts of the globe. The following table and
diagram may lead to some, if sketchy, answers.
We notice the completely identical pattern of the variations of the Lake
Constance level (Fig. 2, scale 1: 100) and those of the precipitation in the
Alps, the water levels of the Baltic Sea, the Caspian Sea etc. Even at Barnaul
and Nertschinsk does the amount of rainfall show a significant decline-a
fact that was pointed out by Woeiko:f6° in 1870 for the years 1845-65,
followed by an equally strong increase in 1871 and 1866 respectively. These
distant areas of Inner Asia participate in Europe's climate change. The same
is the case with tropical Africa as indicated by the levels of the Nile's flood
waters changing from lustrum to lustrum with a minimum around 1835, a
maximum around 1846/50 and a second less pronounced minimum around
1856160, followed again by another increase. India, as is evident from rain
observations at Madras, shows an even closer concurrence with this
variation. Here, the amplitude is more than 0.5 meters in rainfall; and the
rainfall of the driest lustrums in 1831-35 compares to the three wettest lustra
in 1816-20, 1846-50 and 1871-75 at a ratio of2 to 3.
In the New World, climatic variations reveal themselves in the amount of
water discharged in the Mississippi delta as well as the variations in the
water level of the five Great Lakes-here represented by Lake Michigan.
Despite some discrepancies with its closest neighbours, the Great Salt Lake
is also part of this variation. By 1856 the latter reached an, if somewhat low,
maximum followed by a drop of the sea level by 1.5 meters in 1860. Starting
in the early Sixties the lake increased by a full 3 meters. The surprising
result is that all of the countries of the Northern hemisphere61 are at the
present time experiencing the same secular climate variations: a fairly
dry period around 1830, a wet period around 1850 and a second dry
period around 1860, followed again by a wet period (around 1880?).
59 Danube, Rainfall (South of Russia).
60 Wild's Repertorium fUr Meteorologie I 2, p. 199.
61 and, according to my latest research, I may include the Southern hemisphere.
nico.stehr@zu.de
FLUCTUATIONS OF WATER LEVELS 61
We are dealing with atmospheric variations or climatic changes of a
grander scale than those eleven-year ones which some tried to link to sun
spot activity. They are grander because of their larger amplitude as well as
their greater length-variations in which all hydrographical phenomena on
earth participate: glaciers, rivers, lakes, and the relatively enclosed
inlets-all of which grow at the same time and recede again at the same
time.
1166/
70
~
lil::>
Ul
..:
;;;;
..:
..:
j,!
'"::<
Ul
..:
76180 81185 :116190 911'95 16120 2lnS 26130 11/)5 ]6140 S6f(IO 61(65 66170 71nS 761SO
Figure 2.2.
nico.stehr@zu.de
0\
N
Europe Asia Africa North America
Water level of Rain Fall at Water volume Water level of
Lake Nile of the Michigan Great Salt
Constance! Barnaue Nertschinsk22 Madras3 Level4 Mississippi5s Lake6 Lake6
m mm mm mm m cbkm cm cm
1816/20
1816120 +4 1458
1821/25
1821125 +0 1133 619'
1826/30 -2 1303
l303 6.80 527
1831/35
1831135 -14* 914* 6.44*
6.44· 523 tr1
1836/40 -3 1311
l311 6,55
6.55 479·
479* Cl
d
1841145
1841/45 +1 282 514 1262 6.97 659 C
1846/50 +1 291 431 1486 7.30 608 0
>-
1851/55
1851155 +10 221 421 1262 7.13 503 79 ~ES
IJ:)
IJj
1856/60 -17*
-17· 203 351 1174 6.92* 523 84 52* :;;0
i'O
nico.stehr@zu.de
1861165
1861/65 -16 194 325*
325" 1112* 7.20 61 88 C,
C:
(1
n
1866/70
1866170 +5 173* 432 1130 7.39 45·
45'" 274
1871/75 +5 258 408 1432 7.52 483 48 311 1itr1tT:1~
1876/80 +28 369 430 534 :;;0
i'O
1 Beitriigezur Hydrographie des Grossherzogthums Baden, I. Issue, 1884; Honsell, Der Bodensee, Atlas Table. IV.
2 Annalen des physikalischen Centralobservatoriums zu St. Petersburg.
3 Fritz, Beziehungen der Sonnenfiecken
Sonnenflecken und den magnetischen und meteorologischen Erscheinungen der Erde [Relation of Sun Spots to
Magnetic and Meteorological Phenomena on Earth], Haarlem 1878, p. 129.
4 Fritz in Zeitschrift flir Meteorologie, 1878, p. 363.
ff.
5 v. d. Groeben, op. cit., 1884, p. 1 if.
6 Woeikof, op. cit., 1881, p. 288.
Chapter 3
How Constant is Today's Climate?*
The question of climatic change in historical times has always been of great
interest as it has far reaching practical consequences. Once the evidence is
before us that climatic changes took place on our globe within man's sight,
we have to conclude that changes will continue to happen in the near future.
Such changes, however, will not pass without the profoundest impact on the
life activities of people. Thus, to a certain extent, the general answer to this
question is a prognosis of man's future fate and that of his accomplishments.
This question has been answered in many different ways, and there are
few conceivable cases of climate change whose occurrence has not been
claimed for either smaller or larger areas.
There can be no doubt that the climate of the geological past was
different from today's. Fossils found in the polar zone give us the same clear
evidence of a once tropical vegetation during the earliest periods of Earth's
history, such as the Cretaceous Period, as the moraines of the diluvial
glaciers in the moderate zones and even in certain tropical mountains.
Climate has changed from the Cretaceous Period up to the ice age and since
the ice age; but can we prove this latter change in historical times? The
answer has been a multiple "yes". In 1882, Whitney, for instance, tried to
show that the whole planet experienced a general droughtY A similar
climate change is claimed by Theobald Fischer when he is talking about the
advancing deserts in the Mediterranean. 63
However, more often than these geological changes of the climate, as one
is inclined to call them, during historical times one has claimed to have
• In wie weit ist das heutige Klima konstant?, Verhandlungen des VIII. Deutschen
Geographentages, Berlin, 1889.
62 Whitney, Climatic changes in later geological times, Memoirs of the Museum of Compara-
tive Zoology at Harvard College. Vol. VII. Cambridge, 1882.
63 Th. Fischer, Beitriige zur physischen Geographie der Mittelmeerliinder [Essays of Physical
Geography of the Mediterranean], Leipzig 1887, p. 25 ff. and Studien tiber das Klima der
Mittelmeerliinder [Studies of the Climate of Mediterranean Countries]. Additional issue
no. 58 to Petermann's Mitteilungen. Gotha 1879, p. 41.
63
nico.stehr@zu.de
64 EDUARD BRUCKNER
observed local changes restricted to a particular area and caused by human
actions. Numerous researchers held for instance the opinion that the clear-
cutting of forests may diminish the amount of rainfall and the water supply
in springs and rivers, while it increases with reforestation. On the other hand
Americans living in the dry zones of the western United States firmly
believe that extensive cultivation has increased the rainfall of the region.
The many relevant hypotheses and theories which in this or a similar
fashion are defending an ongoing unidirectional change of the climate are
contradicted by an equal number of findings denying a change of this kind in
historical times. It is notable that it is usually the meteorological science that
rejects the idea of a continuous change of climate, while geologist~,
geographers and hydrographers have frequently supported it. For
meteorologists, the constancy of the climate is, to a certain extent, axiomatic.
It appears quite inexplicable how such a conflict of opinions could
develop.
The question entered into a new phase when a one-way change was no
longer investigated and meteorological data material were instead examined
with respect to secular weather variations. This was instigated by the
mysterious fluctuations of the alpine glaciers whose only cause had to be the
meteorological changes.
Forel64 and Richter65 were the first to attempt to prove this theory
statistically. It gained general acceptance through the research by C. Lang in
1885 66 which, based on more extensive data about the entire region of the
Alps, identified alternating long-term periods of fairly cool and moist and
relatively warm and dry weather in relation to the glacial changes.
This was the situation two years ago when I discovered analogous
weather variations with a completely different approach and for very
different areas and presented my first findings at the German Meteorological
Society's meeting in Karlsruhe. 67 Hydrographical research made me aware of
unusual fluctuations of the water levels of the Baltic Sea, the Caspian Sea
and the Black Sea, all of which showed a rhythmic pattern, as Swarowsky
64 Forel in Archives des sciences phys. et naturelles, 1881, No.5.
65 Ed. 'Richter, Der Obersulzbachgletscher [The Obersulzbach Glacier], Zeitschrift des
Deutschen und Osterreichischen Alpenvereins, 1883, p. 75 fT.
66 C. Lang, Der siikulare Verlauf der Witterung als Ursache der Gletscherschwankungen in
den Alpen [The Secular Weather Changes as a Source of Glacial Variations in the Alps],
Zeitschrift der Osterreichischen Gesellschaft fUr Meteorologie, 1885, p. 443.
67 Ed. Bruckner. Die Schwankungen des Wasserstandes im Kaspischen Meer, im Schwarzen
Meer und in der Ostsee [Fluctuations of Water Levels in the Caspian Sea, the Black Sea
and the Baltic Sea Relative to Weather], Lecture, Annalen der Hydrographie 20. 1888,
February issue. A short version of the lecture was published in the Meteorologische
Zeitschrift, 1887, June issue, p. 232. [Reprinted as Chapter 2 in this publication.]
nico.stehr@zu.de
HOW CONSTANT IS TODAY'S CLIMATE 65
had discovered it at the Neusiedler Lake68 with a distinct similarity to the
rhythmic variations of glaciers. At all locations and at the same time, periods
of generally high water levels alternated with those of fairly low levels. At
the Caspian Sea it was obvious to seek the cause in a change of the amount
of water supplied by its rivers and in the lake's evaporation. An assessment
of water level observations at the Volga as well as observations of
precipitation carried out by a number of meteorological stations in the
Russian Empire changed assumption into fact. It turned out that the same
variations of rainfall Lang had identified for the Alps were also occurring in
the enormous drainage area of the Caspian Sea. Furthermore, the areas of the
Baltic Sea and the Black Sea were subject to the same rainfall variations as
well, and the peculiar long-term level variations occurring at these seas are
partly a result of the water supplied by their tributaries which fluctuates
along with the rainfall. The average water levels of the Weichsel, Oder,
Elbe, Weser, Rhine, Danube, and even the Seine, all of them reflect clearly
the variations of rainfall with the same long-term periodicity. In short, we
find these secular variations of weather conditions all over Europe. In testing
some of the meteorological and hydrographical data available we learnt that
more or less all countries of the northern hemisphere participate in these
variations: their universal occurrence and length enable us to identify them
as climate variations. I was very pleased when in the fall of 1888, based on
his research on the level variations of a number of lakes, Sieget'9 confirmed
the majority of my findings.
Meanwhile I was able to gather additional data relevant to the subject and
to extend the research to the southern hemisphere as well.
The initial results were primarily achieved on the basis of hydrographical
phenomena and consequently of a qualitative nature. In examining the
meteorological data, it was now necessary to also give a quantitative account
of the climate variations.
Today we have standardised observations from nearly 600 meteorologi-
cal and hydrographical stations adding up to a total of 30,000 years of
observations, and with the help of these data we will already be in a position
to gain a clear picture of climate fluctuations experienced on our globe. I
68 A. Swarowsky, Die Schwankungen des Neusiedler Sees [Fluctuations of the Neusiedler
Lake]. Bericht tiber das XII. Vereinsjahr des Vereins der Geographen der Universitlit
Wien, 1886, p. 15.
69 R. Sieger, Die Schwankungen der hocharmenischen Seen seit 1800 In Vergleichung mit
einigen verwandten Erscheinungen [Variations of the Upper Armenian Lakes since 1800
in comparison with some related phenomena], Mitteilungen der K.K. Geographischen
Gesellschaft, Wien, 1888.
nico.stehr@zu.de
66 EDUARD BRUCKNER
may be permitted to briefly submit some of the main results from my nearly
completed research. 70
Already, the characteristic variations of hydrographical phenomena such
as measurements taken at glaciers, lakes and rivers, made it seem most likely
that climatic changes would be particularly evident during rainfall. In fact,
this proved to be the case.
The following diagrams (Figs. 1 and 2) show the graphic curves of the
rainfall variations. The first series of curves reflects the entire northern
hemisphere, the second series spans from the northern hemisphere through
the tropics to the southern hemisphere. Each curve represents the variation of
a large territory based on the average measurements from many stations. The
consulted stations are:
1. Scotland (Arbroath, Laurick Castle, Loch Leven Sluice, Northesk
Reservoir, Glencrose, Swanton, Fernielaw, Edingburgh, Inveresk,
Haddington, Culloden, Sandwich, Arrdaroach, Castle Toward, Cameron
House, and Bothwell Castle). 16 Stations.
2. England (Chillgrove, Nash Mills, Oxford, Exeter, Orleton, Podehale,
Boston, Bolton and Kendal). 9 Stations.
3. Northern France (Rauen, Paris, Vendome, Pannetiere, La Collancelle,
Clamecy, Avallon, Laroche, Montbard, Poully, and Dijon). 11 Stations.
4. Northern Germany (Kleve, Trier, Koln, Boppard, Giitersloh, Frankfurt
(Main), GieBen, Bremen, Kiel, Heiligenstadt, Torgau, Dresden, Stettin,
Berlin, Kiistrin, Frankfurt (Oder), Posen, Gorlitz, Breslau, Konigsberg
and Tilsit). 21 Stations.
5. Austria-Hungary (Bodenbach, Praha, Deutschbrod, Lemberg, Krems-
miinster, Klagenfurt, Wien, and Hermannstadt). 8 Stations.
6. West Russia (Helsingfors, St. Petersburg, Riga, Warsaw, Moscow, and
Kiev). 6 Stations.
7. East Russia (Lugan, Ssimferopol, Astrakhan, Baku, Tiflis, Bogoslows,
Jekatherinenburg, and Slatoust). 8 Stations.
8. West Sibiria (Barnaul). 1 Station.
9. East of Sibiria (Nertschinsk (metallurgical plant), Nikolajewsk (Amur),
and Peking). 3 Stations.
10. United States, North America, Interior, (Toronto/Ontario, Milwau-
kee/Wisconsin., Detroit/Michigan, Madison/Iowa, Steubenville/Ohio,
Marietta/Ohio, Cincinnati/Ohio, Leavenworth/Kansas, and St.
Louis/Mississippi). 9 Stations.
11. Central Italy (Parma, Modena, Bologna, Genoa, Florence, Sienna, and
Rome). 7 Stations.
70 This inquiry will be published by early 1890 in Penck's Geographischen Abhandlungen
(Wien, Holzel) under the title Klimawandel seit 1700 [Climate Changes since 1700]. It
will include all the research material in detail which served as basis for the curves.
nico.stehr@zu.de
HOW CONSTANT IS TODAY'S CLIMATE 67
12. India (Madras, Calcutta, Jablapur, and Bombay). 4 Stations.
13. Mauritius (S. Louis, Alfred-Observatory). 1 Station.
14. Australia (Adelaide, Bathers, Bukelong, Deniliquin, Goulburn, Mel-
bourne, and Sydney). 7 Stations.
The graphical charts are based on observations from a total of 111
meteorological stations. After averages for each country were derived from
these stations, the lustra averages of rainfall were determined for each station
and expressed in percentages of the thirty-year mean of 1851-80; the averages
of each country based on data from several stations were then smoothed
according to the following formula: (a + 2b + c)/4, i.e., the first and last lustrum
according to the formula (2a + b) 13 and (a + 2b) 13.
The curves of the chart give a clearer picture of the data trend.
A rise and fall of the curve by one increment refers to an increase and
decrease of rain by 5%. The distance between the top and the bottom of each
curve shows the amplitude of the variation, in relative not in absolute terms.
The wider the gap, the greater the difference between the maximum and
minimum amount of rainfall.
We see a surprising similarity in the shape of the curves with an overall
decrease towards the year 1860 and an increase towards 1850 and 1880, from
the west coast of the old world to the east coast as well as in the interior of
North America, from Germany via Italy and India up to Australia. It is evident
that more or less all of the countries around the globe go through a rainy period
and a dry period at the same time. In the current century the rainfall maxima are
grouped around the years 1815, 1850, and 1880, the minima around 1830 and
1860.
In absolute terms of course, the epochs do not coincide perfectly; in some
areas the minimum precipitation occurs in 1856-60, in others around 1861-65,
in one case even not until 1866-70, and the maximum shifts accordingly. In
none of our cases, however, does a minimum coincide with a maximum. No
minimum occurs in the years 1841-55 and 1871-85 and no maximum in the
years 1825-40 and 1856-70. Thus, too much rain in one region does not mean
too little rain in another; a compensative process does not take place on the
global continents concerned. The slight deviations derived from the epochal
mean are irregular and mostly confined to smaller areas. However, this cannot
be otherwise. Just as despite a clearly defined annual periodicity the maximum
of a meteorological element, such as temperature, may occur earlier or later in
the year depending on the actual weather conditions, it is the case here.
nico.stehr@zu.de
68 EDUARD BRUCKNER
1831/35 36/40 41145 46/50 51155 56/60 61165 6&70 71n5 76/80 81185
I..- ~ ['.
Scotland I I I I I I t q:: 1 I I I I I Id 70
1 108
---
~ r-
-
~,
England
r- iooo. .....
., - ...... r- 102
North France
~ I"""- i'oo...
.....
r-....
-..... ..... ~
~
.....
f-'
106
110
North Gennany ." ".
~~ :--.. ~ i'
Austria-Hungary 1/ ....... ..... ~ 107
" .i '
V ....... 108
-
10'"
Western Russia
,... ..... .... ~ / 167
Eastern Russia i '
..... ~
/ i'
"" '" ...... ~
~
:/
I
II
Western Siberia .... ....... J 281
I
1/ ".,
''''
II
i.-'"
~
~
, J. ~
,
Eastern Siberia If ,,~
/rJ
'/
\' / I
, II
- -.....--
I
I' ~
106
./
..... .....
Continental ~ ~
J"
1-- ~
'"
USA
Figure 3.1. Secular Variations ofRainfall: 10 Sectionfrom West to East
Also the relative intensity of the maxima and minima is not the same
everywhere: in Australia the maximum around 1850 is more pronounced than
the one in the 70s; in a number of regions the two maxima are equally intense,
while in the majority of cases the maximum around 1880 is stronger than the
one around 1850.71
71 Some of these instances can most likely be attributed to the improved methods of
measuring the precipitation, particularly during the winter, as the general upward trend in
the curves for Russia, Siberia, and Centr~l Italy seems to indicate.
nico.stehr@zu.de
HOW CONSTANT IS TODAY'S CLIMATE 69
1831135 36/40 41145 46/50 51/55 56/60 61/65 66170 71175 76/80 81/85
North Germany -I-- 106
,......
V ...... """
-
~
,......
~
"-
-
L /
104
- Central Italy ~
I"..ooo~~ ~ .... ~
~
-
.". ~ """ ""'" 107
"
r--. """ . / ~
~
"India
........ ~ ",
/ ""
..... i""'"
,~ -1""-0 103
V - ~ Mauritius/
~
.,
Australia V ~
'"
~
f-
"- ........ ~ / ."\
"\
'- 91
Figure 3.2. Secular Variation a/Rainfall: Section from North to South
However, certain areas seem to be the exception to the rule. There is
southern Italy and Sicily, as well as southern Spain, there are the lower valleys
of the Indus and the Ganges, furthermore the eastern United States, where a
maximum rainfall occurs in the 60s, during a time, that is, when other countries
h51ve little rain. Iceland seems to be an analogous case. Also Scotland
displaying variations that are in part extremely vague is an exception while
England adheres to the rule. These exceptional cases are, as far as we are able
to tell from the present state of meteorological and hydrological observation
methods, quite insignificant in comparison with the bulk of the land masses
exposed to the variations.
Another pattern emerges clearly from the curves: the variation is more
distinct as it advances further into the interior of the continental land mass. In
Scotland it is vague. In Germany it is distinct and the rainfall of the driest
lustrum around 1860 compares to the amount of the lustrum with the most
rainfall in 1880 at a ratio of 1: 1.09; in the eastern parts of European Russia this
ratio is 1:1.24, and in West Siberia even 1:2.26, where more than twice the
amount of rain fell during the wet 5 years of 1881-85 than during the dry years
of 1861-65. In East Siberia the ratio falls again to 1: 1.36. In view of this fact,
and especially when considering the geographical location of some of the
exceptional regions along the coasts of the Atlantic Ocean, it seems logical to
conclude that the rhythmic pattern of the rainfall variation has to do with the
landmass of continents and that compensation might take place over parts of
the open ocean.
nico.stehr@zu.de
70 EDUARD BRUCKNER
Rain is not the only meteorological element displaying this kind of rhythmic
variation. Similar ones have been identified for the temperature in a two-fold
approach: one based on records about the length of time of the rivers' ice cover
during winter and the other based on temperature measurements. The following
chart, Figure 3, will illustrate these temperature variations. The resulting curve
is based on lustra averages of the mean values reported by Koppen72 and
determined by those stations in Europe and New England which carried out
observations even prior to 1820. An increase of the curve by one equals a
temperature increase of 0.05 degrees Celsius.
1786/90 91195 96100 1801105 06110 1lI15 16120 21125 26130 31135 36140 41145 46150 51155 56160 61165 66170
0.25 , , 0.25
0.20
0.15
V
r-
"'"" f\ I 1'\" \ 1/ """
0.20
0.15
0.10 0.10
0.05 / \. I \ 1/ 0.05
0.00 1/ I- ...... i"""
0.00
-0.05 1\ J l\ IL -0.05
I
"I" "- L
-010 -0.10
-0.15 \. I -0.15
-0.20 / -0.20
-0.25
1\ / -0.25
-0.30 ' , -0.30
1786/90 91195 96100 1801105 06110 1lI15 16/20 21125 26/30 31135 36140 41145 46/50 51155 56/60 61165 66170
1676180 81185 86190 91195 96100 1701105 06110 11115 16/20 21125 26130 31/35 36140 41/45 46150 51/55 56/60 61/65 66170 11m 76180
-50 -5.0
40
j 4.0
-3.0 -30
,
·2.0
I"\, -2.0
-1.0
00
.......... r-... -1.0
0.0
1/ I'
"
10 10
20
i""" I- 2.0
30
..... 1/ "\. \.
-"1 r-... 30
;~ I I I I I I I I I I I I I I I I I I I I I I I I I CH 111 I I I I NW I:~
1676180 81185 86190 91195 96100 1701105 06110 11115 16/20 21125 26130 31/35 36/40 41/45 46150 51/55 56/60 61/65 66170 11m 76/80
1776180 81185 86190 91195 96100 1801105 06110 11115 16/20 21125 26130 31135 36/40 41/45 46150 51/55 56/60 61165 66170 11m 76/80
-5.0 5.0
40
""1- 40
-30 I 1\ 3.0
-20 2.0
-10
I- 1\ 10
0.0
V If
V ,\ 0.0
1.0
-t-- l/I"""!- 1.0
20 2.0
3.0
I\. /' I' I"" / \. 3.0
1/
~~IIIIIIIIIIIIIIm1IIIIIIIIIIIIIIIIIIIIIII~~
1776180 81/85 86190 91195 96100 1801105 06110 11115 16/20 21125 26130 31/35 36/40 41/45 46150-51/55 56/60 61165 66170 11m 76180
Figure 3.3. Secular Temperature Variations
72 Koppen in Zeitschrift der Osterreichischen Gesellschaft flir Meteorologie, 1874, p. 266.
nico.stehr@zu.de
HOW CONSTANT IS TODAY'S CLIMATE 71
One general finding is obvious: the rhythmic pattern of temperature
variations is such that wet periods are at the same time cool, and dry periods
warm. For the current century we encounter, for instance, a generally cold
period around 1815 in Europe and New England, a warm period around 1825, a
cool period again around 1840, the years of the 60s are warm again followed by
the last cool period at the end of the 70s.
As a next step, the question arises whether these climatic variations can also
be identified for earlier centuries. One succeeded by relying on a number of
different data. Of course, reliable meteorological observations do not date back
very far into previous centuries; yet other valuable data were available, based
partly on hydrographical and partly on plant-phenomenological factors. Among
them are, for instance, the water level variations of the Caspian Sea recorded
since 1685; there also are observations of the freeze-up and break-up of the ice
on several of the Russian rivers, which with few interruptions go back to the
beginning of the previous century. Foremost however, there are the dates of the
beginning of the grape harvests in the wine areas of France, Southwest
Germany and Switzerland recorded over many years and starting as early as
1400. Not until the year 1550 however is the number of recording stations large
enough to allow cross-checking and to obtain an average.
It is common knowledge that grapes tend to mature late during cool, wet
years and early in dry, warm years. As we have evidence that the changes in
the harvesting dates until the middle of the previous century coincided with
the annual changes in temperature and rainfall, they may be considered the
ultimate indicators of climatic variations.
The lustra averages of the harvesting dates from 30 stations in the three
countries mentioned were determined (Verdun, Argenteuil, Foug, Loches, Les
Riceys, Coupignon, Denainvilliers, Auxerre, Vendome, Vesoul, Dijon, Beaune,
Lois-Le-Soulier, Renne, Pichon-Longueville, the Medoc, Tain, Castres,
Stuttgart, Kiirnbach, Altstetten, Veytaux, Vevey, Lausanne, Lavaux, Aubonne,
Rolle, and Pully near Morges).73
For each station, these lustra averages were expressed in deviations from the
30-year mean of 1851-80 and based on the stations' averages. The total
average was then determined for each lustrum. These values have been
adjusted as before and are shown in the curve above allowing us to trace the
climatic variations back to 1670. I could even have traced them back to 1550,
though with less accuracy, because as of 1670 the number of recording stations
goes down rapidly. But the sample provided here should suffice. 74 An increase
73 The majority of data series collected by Angot Vendanges en France. Annales du Bureau
Central Meteorologique de France 1883, I. Paris 1885.
74 The curve starts in 1550, and, as all other observational material, will be explained in more
detail by the author in the extensive publication mentioned in footnote 70.
nico.stehr@zu.de
72 EDUARD BRUCKNER
of the curve by one increment equals an earlier start of the grape harvest by one
day, that is an increase in temperature and a decrease in rainfall.
Around the years 1880,1851-55,1816-20,1766-70,1741-45,1696-1700,
and 1671-75 we register cool and wet periods, around the years 1861-65,
1820-30, 1786-90, 1756-60, 1726-30 and 1681-85, warm and dry periods.
The time spans between two maxima are not completely the same: climatic
variations do not take place with the precise periodicity of a certain length of
time and when we talk about a periodicity of 36-37 years it is not more than an
average value.
What causes these characteristic variations of the most basic climatic
elements? The ultimate reason is still a complete mystery. Data about rainfall
are the only ones that can be explained reasonably, and only for the limited
region of Europe. Here, the variations in rainfall go exactly hand in hand with
secular variations of the air pressure. These take place over the North Atlantic
Ocean and over all of Europe in such a way that during rainy periods the
pressure differences are reduced, and increased during dry periods. During
most parts of the year and on average the air flows from the continent to the
ocean; when the continental system is weakened, there is less resistance to the
oceanic air mass that can then move inland causing precipitation. When the
system gains strength, the opposite effect is achieved. Fluctuations of air
pressure are nothing but the results of fluctuations in temperature. An increase
of the latter leads to a sharper thermal contrast between water and land in areas
of moderate and upper latitudes causing high-pressure ridges over land with
less precipitation. Similarly, the fluctuations of the rainfall should, in general,
also be attributable to those fluctuations of temperature, though at present this
has only been proven for Europe. He who can explain the temperature
variations, has found the cause of the entire phenomenon of climatic variations.
Are climate variations so significant that they are of practical impact?
Indeed they are. In dry areas in particular where water is notoriously scarce the
hydrographic conditions change dramatically during periods of climatic
variations. Lakes disappear during dry periods and reappear during wet ones, as
for instance Lake George in New South Whales which in 1820 and again in
1876, and to a lesser extent in 1850, used to be a large lake of 12 to 18
kilometres in length, 10 kilometres in width, and 5 to 8 meters in depth, yet
disappeared completely in the dry periods in between; or some central African
lakes such as the Tshad, Tanganyika and Nyassa, which, according to Sieger,15
at times rise so high that their overflowing waters create an outlet lasting for
some years and which then lose this outlet again when the dry period begins.
Rivers and creeks dry out for a full decade; swamps dry up and reappear in the
75 R. Sieger, Schwankungen der innerafrikanischen Seen [Fluctuations a/Inner Africa Lakes].
Bericht tiber das XIII Vereinsjahr des Vereins der Geographen an der Universitat Wi en,
1888.
nico.stehr@zu.de
HOW CONSTANT IS TODA Y'S CLIMATE 73
next wet period. I could give many more examples in Africa and Australia, if
time would permit it.
Consequently climatic variations deeply affect human life. River navigation
to a great extent depends on the amount of water in the river bed which
determines its depth. In those dry years around 1830 and 1860 shipping
problems increased and soon a lot of speculation began about the possible
cause of the lower river-water levels. In most cases the increasing practice of
deforestation was found to be the source. Now we know better: it is because of
climatic changes.
Another way in which temperature variations are affecting traffic is through
the length of time of the rivers' freeze-up. For example, during the cold spill
from 1806 to 1820, the Newa and the harbour of St. Petersburg remained
blocked by ice for more than three weeks longer than they did during the warm
period from 1821 to 1835.76 This means that during cold years harbours in a
more westerly location and with shorter closure times handle part of St.
Petersburg's shipping traffic which they lose again during warm periods. Of
course, the navigational period may also change in length from one year to the
next. In that case, however, the following year may make up for the loss of the
previous year; not so in the case of climate variations when average values have
changed and favourable and unfavourable conditions cannot be reversed so
soon. Thus certain changes in shipping traffic go hand in hand with climate
changes.
Another area that is affected by climatic variations is agriculture. This is of
special significance for the interior of the continents where variations are more
extreme, e.g., Siberia which suffered famines as a result of the droughts in the
60s; or Egypt where the Nile's floodwater levels determine the fertile
productivity of vast areas of agricultural land.
A future climate change may turn out to be quite disastrous for the arid
areas surrounding the Salt Lake in North America. From the early 1860s to
the middle 1870s the water level of the Great Salt Lake rose by 3 m-filling
its tributaries whose waters were used for the irrigation of recently cultivated
fields and grass lands. 77 We may recall the prevailing view that extensive
cultivation of previously desert-like areas has increased rainfall
considerably.
Contrary to this, I would like to point out that the climatic improvement
coincides exactly with a period of increased rainfall experienced all over the
globe and in continental areas in particular. On the other hand, simqar
changes of climatic conditions have occurred in America before, as we were
able to demonstrate through observations of rainfall and river levels in the
76 Based on data collected by Rykatschew (Closure and Reopening of Russian Rivers), see
charts.
77 Compare Gilbert in Powell, Report upon arid Regions, Washington 1879, p. 57 ff.
nico.stehr@zu.de
74 EDUARD BRUCKNER
neighbouring Mississippi area for the current century. They are the same
variations as the ones which in Europe could be traced back to 1550, and by
this large number of verified variations we are convinced to believe that they
will most likely continue to occur in future. It seems to me highly probable
that the climate improvement in the Salt Lake area that began in the 60s will
now be followed by an arid deterioration in climate. Early signs of this
development can be seen in the recent arid years which in fact caused the
Salt Lake to fall: in 1888 the lake's water level had again returned to its low
marks of 1864. 78 If this hypothesis holds true, then this area will be unable to
avoid a major economic crisis in the very near future; because agricultural
land which was cultivated during the years 1870-80 would soon stop to
produce because of aridity. We would see the same development here, as we
have seen in Egypt and Siberia, namely that the size of cultivable land
changes along with climatic variations.
However, time is running out, let us conclude!
We were able to identify general variations of the climate; and I have
tried to outline the situation briefly. At first glance it may seem unusual that
these variations had escaped scientific scrutiny to this day. However, there
has been speculation about them before: every now and then, based on
unusual sightings at some source or body of water the view emerged in
publications that the climate of certain locations, their rainfall in particular,
was probably subject to periodic changes, as a.o. by Hann for the area of the
Caspian Sea,19 Schweinfurth for parts of the Mediterranean countries,80 and
most of all Fritz for many areas on the globe. 81 However, for the Alps the
only meteorological evidence was provided by Forel, Richter and Lang. And
the universal occurrence of the phenomenon, its global importance and
simultaneous course could by strictly meteorological standards not be
verified to this date, before a large number of meteorological stations had
recorded the dry period of the 60s and the wet period around 1880.
These general climatic variations are the key to the prevalent great
confusion about the issue of climate changes, which we attempted to
describe at the beginning: they are the explanation for the fact that such
78 According to a handwritten curve by G. K. Gilbert which I gratefully received from Dr. R.
Sieger.
79 Hann in Zeitschrift der Osterreichischen Gesellschaft flir Meteorologie Vol. II, 1867.
80 Schweinfurth in the preamble to Biideker's Egypten I. 1877, p. 79.
81 Fritz in Petermann's Mittheilungen 1880, p. 245 ff.
nico.stehr@zu.de
HOW CONSTANT IS TODAY'S CLIMATE 75
contradictory opinions could exist side by side: the climate changes over
time first in one direction and, then again, in another-the climate fluctuates
and with it fluctuate rivers, lakes, and glaciers.
nico.stehr@zu.de
Chapter 4
Climate Change Since 1700'
4.1 THE CURRENT STATUS OF THE INQUIRY
INTO CLIMATE CHANGES"
Definition of the Climate Concept. I. The climate in the geological past.
Lyell's view. Heer's homogenous climate of the pre-Tertiary age. The cooling
of global climate during the Tertiary (Heer and Engler). Neumayr questions
the homogenous climate of the Jurassic, Cretaceous, and Carboniferous
periods. The Carboniferous Ice Age. Nathorst and Neumayr advocate a shift of
the earth axis during the Tertiary period. The Climate of the Diluvial Age. -
II. The Question of climate changes in historical times. a) Changes of
rainfall. Statements of general changes of precipitation around the globe.
'Desiccation' Theory by Whitney. Th. Fischer. Critique of research material.
Partsch contra Fischer. Purported local variations of rainfall in temperate
climate zones and in the tropics: Decrease of water resources from
deforestation or increase from reforestation. Water level reductions of running
water in cultivated areas: Wex. Comparative observations on rainfall
unacceptable for deciding whether forest has any influence on same. The
experimental research of Blanford and Gannet. Rainfall reductions unverified
by observations. There is no continuous decrease of water resources in
cultivated regions. Increase of rainfall due to deforestation in Australia.
Alleged influence of cultivation of land on rainfall in the interior regions of
North America. - b) Temperature changes. General cooling of the climate in
the Northern Hemisphere, sometimes suggested-sometimes refuted. Constant
temperature in historical times: Ideler, L. Dufour. Changes of wind conditions.
Summary: A continuous thread is missing in the maze of hypotheses about
climate changes. - III. Meteorological Cycles. Hypotheses of recurrent
seasonal weather periods within a certain time span. Multi-year periods of cold
cinters: Krafft, Renou, Koppen. Influence of sunspots on meteorological
• Klimaschwankungen seit 1700, Wien; E.D. Holzel, 1890.
.. [Chapter One: Der gegenwartige Stand der Frage nach den Klimaanderungen.]
77
nico.stehr@zu.de
78 EDUARD BRUCKNER
factors: temperature, rainfall, air pressure, etc. Glacial variations in connection
with secular variations of the weather: v. Sonklar, Forel, Richter, Lang.
Just as wind and weather change daily, as rain follows sunshine and
sunshine follows rain, dry and wet, cold and warm years alternate in the
same manner. If for a particular location the annual average temperature is
determined on the basis of temperature readings taken at certain times daily
and compared to averages of the previous or next years, it is not unusual to
find differences of 2° to 3°C or even more. When these temperature
variations from one year to the next are depicted graphically, the result will
be an irregularly shaped 'zigzag' line. However, its shape is not completely
arbitrary but shows distinct deflective clusters around an imaginary centre
line. The latter represents the average annual temperature of that particular
location, whereas the irregular curve represents the weather of consecutive
years. "Climate, as we understand it, is an aggregate of meteorological
phenomena defining the average atmospheric condition at any particular
location on the globe. What we call 'weather' is only one phase, one single
event in a series of climatic phenomena, the entire more or less constant
(annual) cycle which determines the climate of a particular location. Climate
is the summation of weather events over a longer or shorter period of time as
they usually tend to occur at that specific time of the year."82 Hence, to us
weather is linked to unstableness or variability, climate to stableness,
changing from location to location only, but not from one time to another.
People's awareness of the invariability of the climate is deep-rooted and is
expressed in the firm conviction that unusual weather conditions in one
season or year will have to be compensated for in the next.
Nonetheless, even climate did not remain the same over long periods of
time. Numerous hypotheses and theories were formed about the change of
climate in the past and, naturally, occupied the more or less keen interest in
more than one quarter, because a clear confirmation of climate changes in
the past immediately raised the possibility of future change which in tum
could.not be without decisive effects on the economic situation of future
generations. It is this practical aspect, to be sure, that may explain why there
are so many different hypotheses that hardly an incident of climate change
imaginable remains which has not found an advocate. In this confusion of
contradicting opinions, often with little foundation, it seems no wonder if
presently it almost violates conventional rules among meteorologists to even
deal with this question of climate- change, let alone to add a new hypothesis
to the existing ones.
It cannot be our intention to compile ~ll these different opinions on
climate change; such an even remotely complete collection would easily fill
82 Hann: Handbuch der Klimatologie [Handbook o/Climatology] , Stuttgart, 1883, p. 1.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 79
many volumes. But it may be acceptable to roughly sketch the difference of
opinion prevailing to this day by categorising these legions of hypotheses
and theories according to the most common viewpoints.
4.1.1 Climate of the Geological Past
Our knowledge that the climate of the geological past was different from our
climate today is as old as the science of geology. However, there are very
few areas where speculation ran as far ahead of establishing facts as
concerns this issue. Even before a mere fraction of data on this pertinent
subject became known, many were prepared to offer theories on climate
variations in geological prehistoric times, which were believed to have been
recognised as such. As a result, from the beginning of this century to the
present day one hypothesis followed the next, and while most were of no
value, others prepared the future for a more realistic basis of theories and
speculative discussions of different explanations for climate changes. Of all
these, we can discard the purely speculative theories since not one of them
comes even close to satisfying scientific research standards. We will revert
to a brief description of those opinions, which were uttered about the climate
of different geological formations based on facts and which remain valid
even to this day.
It had to be discovered quite early that the prehistoric environment of a
particular place, remnants of which were found in the form of fossils in
geological formations, was exposed to a different climate than it is now. The
idea of a changing climate seemed therefore plausible. Two different routes
could be taken to find out more about the prehistoric climate: Either one
compared organic forms of different geological formations with those of the
present and tried to identify climate changes over time by looking at the
characteristic transformations of those forms; or one focused on a
comparison of different flora and fauna from one and the same prehistoric
period and strived to establish the presence or absence of climatic patterns
for that time period, as they can be observed today from pole to equator.
Both methods have been used.
The first approach is the older of the two and has been chosen quite
frequently since the beginning of this century. It first led to the theory of a
gradual progressive cooling of the global climate. Lyell, however, came to a
slightly different conclusion. 83 Although he, too, was led to believe that past
climates were generally warmer; but this warmer climate had temporary
periods of lower temperatures not only during the Diluvium but also in the
83 Lyell: Principles a/Geology, 10th ed., Chap. X-XIII.
nico.stehr@zu.de
80 EDUARD BRUCKNER
Miocene, in the Eocene and in the Permian. And just as his predecessors, he
hardly touched the issue of climate zones.
Heer's approach was quite different: he carefully compared plant
remnants of the same time period originating from different places of the
globe. To him, his studies of the fossil flora seemed to reveal the fact that
over long periods of time the global climate had been homogeneous and
tropical, which is inconsistent with the solar climate of today and its climate
zones between equator and pole. Modem arctic research and the journeys of
NordenskjOld in particular, who travelled to the polar regions with the
intention to examine prehistoric climate conditions, confirmed the
assumption that findings from the hot and temperate zones could be applied
to the far North as well.
This homogeneity of the climate could, of course, only be claimed for
some eras of our geological past, mainly for the Carboniferous Period.
Nares' polar expedition 84 discovered the same abundant flora in the lower
carbon strata below the 74° and 76° latitude, which grew in temperate
climate zones in the Old and New World at that period, while they
discovered tropical corals and cephalopods in the dolomite at the northern
coast of Grinnell Island (83 ° latitude). 85 Oswald Heer describes carboni sed
plant species and types discovered on Bear Island and Spitzbergen which
are identical with those found in Europe. 86 Some parts of the Mesozoic era
seemed to show the same results. Nathorst, for example, introduced us to
the same Jurassic flora in Spitzbergen and in India,87 and Nathorst and
Heet8 showed us how deciduous trees appeared at the same time in
Northern Greenland, Europe and North America during the Jurassic Period.
According to Heer and Engler,89 not until the Tertiary Period is the
homogenous global climate interrupted and do today's climate zones
evolve starting from the pole. Both, flora and fauna clearly changed with
more advancing cooler temperatures. Heer's standard groundbreaking work
has shown that in Switzerland plant life grew in the same consecutive
84 G.S. Nares: A Narrative of a Voyage to the Polar Sea during 1875-76 in H. M Ships Alert
and Discovery, London, 1878. Volume II, p. 331 f. cit. by Whitney.
85 O. Reer: Flora Fossilis Arctica. Vol V. Part I, p. 17.
86 Reer: op. cit. above, Volume II Chapter I, Volume III. Chapter I and Vol IV. Chapter 1.
87 Nathorst: Polarforskningens bidrag till forntidens viixtgeograji [The contribution of polar
research to prehistoric plant geography] in A. E. Nordenskjold: Studier ochforskningar
foranledda afmina resor I hOga norden [Studies and research related to my travel in hte
nigh north]. Stockholm 1883.
88 Reer: Flora Fossilis Arctica, Volume I, p. 60 and in other places of this huge publication.
In Volume I, p. 53 op. cit., and Volume VII, p. 226, a short summary of all of Reer's
climatological findings.
89 Engler: Versuch einer Entwicklungsgeschitchte der Pjlanzenweit [A Tentative History of
Plant Life], 2 Volumes, Leipzig 1879 and 1882.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 81
pattern and hence fossilised in the same vertical layers during the tertiary
period as those that are imbedded in layers anywhere between pole and
equator. 9O The tropical Eocene flora is followed by the SUbtropical Miocene
flora which in turn, by the end of the Tertiary Period, changed to a flora
borealis similar to what we have today. Lesquereux confirmed these results
by examining the fossil flora of North America's Rocky Mountains. 91
The advancement of cooler climates from polar regions to lower
latitudes, as indicated by these successive layers of fossilised flora, was
fully confirmed by the discovery of the remains of tertiary plant life within
the polar circle which Nordenskjold in particular used for his studies. Once
again the collected material was put exclusively into Heer's expert hands,
who described the findings in an imposing number of volumes, the "Flora
fossilis arctica". A flora identical in characteristics to the SUbtropical flora
from the Swiss Miocene was discovered in East and West Greenland,
Grinnell-Land, the Lena River delta, and on Spitzbergen.
The question arose whether the arctic tertiary flora belonged indeed to
the same time period as the Swiss Miocene flora or was perhaps older and
had developed at an age when the Swiss Alps still had a tropical climate
and a tropical vegetation. A reliable answer to this question could only be
given by studying the Miocene layers in continuity from the South to the
North, which cannot be done as the continent is cut apart by the ocean.
Heer92 and Engler93 maintained that both, the Miocene flora of the Swiss
Alps and the closely related plant life of the polar region were of the same
age, and assumed, therefore, that the temperature difference between the
polar region and central Europe had been relatively small. Other
researchers however, such as J. H. Gardner and Saporta, saw the arctic
tertiary flora with its subtropical characteristics as an equivalent to the
tropical Eocene and Oligocene flora rather than the subtropical Miocene
flora of central Europe. 94 In this case, the age difference between the
SUbtropical tertiary flora of the arctic polar region and the closely related
SUbtropical Miocene flora of the Swiss and West German regions would
perhaps be comparable in likeness and age to the fossil Miocene flora of
central Europe and today's closely related subtropical vegetation of Japan
and of the Southern United States; and the temperature difference between
central Europe and the polar region would have been greater than Heer had
assumed.
90 Heer: Flora Fossilis Helvetiae.
91 Lesquereux: A Review of the Fossil Flora of North America, Bul!. of the Oeo!. and Oeog.
Survey of the Territories, II. Ser. no. V.
92 Heer: Flora Fossilis Arctica, Volume I p. 73, Volume VII, p. 22.
93 Engler, op. cit. see also Volume I, p. 2.
94 Penck in: Verhandlungen des V. Deutschen Geographentages, Berlin, 1885, p. 33;
Neumayr: Erdgeschichte [Earth's History], Volume II, Leipzig, 1887, p. 510.
nico.stehr@zu.de
82 EDUARD BRUCKNER
In either constellation, the arctic Tertiary can be separated into two
climate zones, the outer one of which, reaching up to the 75° Northern
latitude, contains the Tertiary flora, which shares a number of sUbtropical
species with the Swiss Miocene, while the inner region below the 80°
latitude exhibits vivid similarities to today's boreal flora. 95
According to Heer and Engler we can detect the creation of our present
day solar climate developing in the course of the Tertiary. Cooling can first
be perceived at the pole, which as early as the Oligocene Age may have
taken on such dimensions that the tropical vegetation was forced back to
temperate zones and a subtropical flora had spread in the southern parts of
the polar region, while the cooler climate of the region north of the 75_
latitude only permitted the growth of a boreal flora. This process of cooling
continued during the Miocene and Pliocene periods moving the vegetation
further south in ever widening circles with the regions of tropical
vegetation becoming smaller and smaller. By the end of the Tertiary Period,
in the Pliocene, climate conditions as we know them today had more or less
evolved and a flora very similar to today's grew in central Europe.%
It has been attempted to determine the extent of the cooling of the
climate; it was found to differ greatly from one latitude to another. While
today's tropical regions maintained a tropical vegetation from early on and
no noticeable cooling process took place, the southern part of Central
Europe with its tropical Eocene climate being replaced by the subtropical
Miocene climate and finally in the Pliocene Period by the boreal climate
was found to have suffered a temperature loss of 14°-15°C; at the polar
region, where by the end of the Cretaceous Period plants were still growing
whose descendants are nowadays restricted to the tropics, and where by the
end of the Tertiary the snow and ice cover expanded with some scanty
arctic vegetation still present, temperatures sank by almost 30°C.97
Heer's view, as we have just recounted above, can be summarised as
follows: It is a homogenous climate up to the Carboniferous Age and
possibly in some epochs of the Mesozoic era as well-a cooling process in
the Tertiary Period started in the polar region and progressed from there. 98
The cause of this peculiar cooling of the climate during the Tertiary
Period is still unknown. In any case, the old hypothesis that the cooling of
95 Heer: Flora Fossilis Arctica, Volume I, p. 73, Volume VII, p. 22.
% Geyler und Kinkelin: Oberpliociinjlora aus den Baugruben des Kliirbeckens bei Niederrad
und der Schleuse bei Hochst a: M, Separate print of the Abhandlungen der
Senckenbergischen naturf. Gessellschaft Frankfurt a. M., 1887, p. 43 ff.
97 Also compare Penck: Deutsches Reich [German Empire], Wien, Prag, Leipzig, 1887,
p.107.
98 Penck gives an excellent description of this idea: Die erdgeschichtliche Bedeutung der
Siidpolarforschung [The Geological Significance of the South Polar Research], Verhand-
lungen des V. deutschen Geographentages, Berlin, 1885, p. 25, cont'd.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 83
the earth's inner core had something to do with it, has no merit anymore
today. All serious doubts about the physical methods applied set aside, we
learn from Heer's history of the earth as he deciphered it from the fossil
remains, that a gradual cooling of the climate occurred during the Tertiary
Period and only then, neither before nor after that period. It is highly
unlikely that the cooling process of the terrestrial globe due to a
temperature loss to the universe should have been slow and hardly
noticeable during the long geological periods prior to the Tertiary-only to
accelerate enormously during the relatively short period of the Tertiary.
Although the evidentiary material submitted by said researcher is
substantial and his conclusions may seem logical, doubts have been raised
most recently against his views, and, above all, the theory of a homogenous
climate during the earliest prehistoric periods is nowadays again seriously
being questioned.
There were three categories of arguments favouring a homogenous,
predominantly warm climate on earth during earlier periods. One category
concerns arguments based on the abundant prehistoric vegetation which
had created the huge coal deposits; the second refers to the fact that the
organisms of pre-tertiary periods show a closer analogy and similarity with
today's life forms in the tropics rather than with organisms of colder
regions; the third conclusion assumes that the prehistoric fauna and flora
from very different geographical regions are identical.
As Melchior Neumayr points out, these arguments are partly
inconclusive, incorrect, and erroneous. 99
The first argument is completely wrong, as, by the way, has been known
for a long time. Today it is in the cold regions with rich vegetation that
deposits of carbon substances are formed while in warm regions plant
material is destroyed quickly in the process of decomposition.
The second type of argument is disputed on the ground that the
adaptability of organisms to different climate conditions has been under-
estimated; as Neumayr attempts to show, this change could have happened
over time when competing species evolved with better protection against
the harsh climate than the older species, which were then forced to retreat
to tropical regions. Even today wild growing plants of the tropics can be
cultivated in severer climates as long as they are protected in their fight for
99 M. Neumayr: Uber klimatische Zonen wiihrend der Jura- und Kreidezeit [About the clima-
tic zones during the Jurassic and Cretaceous], Denkschrift Wiener Akademie
Mathematisch-naturwissenschaftliche Classe, Volume XL VII. Also: Die klimatischen
Verhiiltnisse der Vorzeit [Climatic conditions o/the prehistoric age], Schriften des Verein
zur Verbreitung naturwissenschaftlicher Kenntnisse in Wien 1889. Several comments in
his Erdgeschichte [Earth's History] Volume II (Leipzig 1887).
nico.stehr@zu.de
84 EDUARD BRUCKNER
survival from better equipped competitors, that is if they are grown in fields
and beds free of "weeds".
On closer look, the often-emphasised identity between fauna and flora
from different latitudes is not entirely correct either. In contrast, Neumayr
showed that during the Jurassic and Cretaceous Periods the spreading of
certain sea organisms occurred in a zonal structure, which leads him to
conclude that climate zones were present and corresponding in their
distance from the poles to today's climate zones.
According to Neumayr, the flora of the Carboniferous Period too is not
at all absolutely homogenous. On the contrary, some findings can be linked
to different climates, such as the rare occurrence or total lack of Sigillaria
[a lycopodium plant indigenous to the carbon age] in carbon layers from
northern latitudes and the lack of a characteristic carboniferous flora in the
tropics. No doubt the polar regions' annual temperature was higher then
and, above all, winters were milder than today.
The theory of a homogenous climate in the Carboniferous Period may
well suffer the strongest blow, however, if identifying certain carbon
deposits as glacial formations should prove correct. IOO For, in the southern
part of Africa, in Australia, and in India lOl peculiar conglomerates or
composites were observed· in the top layers of carbon, which, embedded in
slate clay and fine meshed sandstone, contain scratched blocks of glacial
appearance; in India and in South Africa lO2 the layers of the deposits even
show [glacial] scratches in certain regions. Therefore everything seems to
join in the effort to insure the presence of extensive glaciers by the end of
the Carboniferous Period. Although in light of the many deceptions, which
even skilled geologists have fallen victim to by mistaking pseudo-glacial
formations for real ones,103 the issue of a carboniferous ice age remains
unsolved at present. The seemingly simultaneous appearance of glacial
formations found in three different locations of some distance apart and in
layers supposedly of the same age might be called striking, if the
100 A detailed explanation to the question is given by W. Waagen: Die carbone Eiszeit [The
carbon ice age] (Annual Report of the k.k. geolog. Reichsanstalt, 1887, pp. 143-192) and
by Feistmantel: Ober die Pflanzen und Kohlen fohrenden Schichten in Indien (bezw.
Asien), Afrika und Australien und die darin vorkommenden glacialen Erscheinungen
[About the Flora- and Carbon-Containing Layers in India, (resp. Asia), Africa and
Australia and their Glacial Indications], Sitzungsbericht der koniglichen bohmischen
Gesellschaft der Wissenschaften, Praha 1887, pp. 1-109.
101 Waagen, op. cit., p. 147.
102 A. Schenck: Die geologische Entwicklung Sudafrikas [The geological development of
South Africa]. Petermann's Mittheilungen, 1888, p. 229 f.; also: Ober
Galcialerscheinungen in Sudafrika [About Glacial Indications in South Africa],
Verhandlungen des Deutschen Geographentages, Berlin, 1889.
103 Compare Penck: Pseudo-glaciale Erscheinungen [Pseudo-Glacial Indications], Ausland,
1884, p. 641 ff.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 85
conclusion that they are all of the same age had not been based in part on
those very same alleged glacial formations. As the case may be, the theory
of a homogenous climate during the Carboniferous Period and the pre-
tertiary periods definitely stands on very shaky ground; according to
Neumayr,I04 "almost everything points against the assumption that the
global temperature was uniformly hot from the equator to the poles at one
time."
The latest research by Nathorst and Neumayr also sheds new light on
the cooling process of the global climate during the Tertiary Period which
according to Heer and Engler, was to have been a continuous,
homogeneous process originating at the poles; it seems, the course of
events was not as simple as assumed.
Neumayr has disputed the continuity of the cooling process. He tries to
show the likelihood that Europe's climate was slightly cooler during the
lower Eocene than during the middle and upper periods, with the highest
temperature in the latter. lOs The beginning of a progressive cooling trend,
which lasted up to the Diluvial Ages, can be dated from that time on.
Nathorst suggests that in Spitzbergen for example the climate was severer
during the upper Jurassic period than during the entire Tertiary Period and
that in Japan no cooling has been observed since the Miocene. I06 Also on
both the Sachalin and Camtschatka peninsulas the tertiary (Miocene) flora
indicates a climate condition only marginally warmer than today's.
Consequently, the cooling during the Tertiary Period was not a global
event, but manifests itself in those parts of Europe, which were first
explored for their fossil flora, and is strongest in Greenland where together
with Grinnell-Land the temperature dropped by almost 30_ C, as mentioned
above. Nathorst points out that especially Japan, a region with no cooling,
and Greenland, a region with a strong cooling, are both located along the
same meridian. The geographical disposition of the Miocene flora from
different climates could be explained by assuming that the pole's location
was below 70° northern latitude and 120° eastern longitude of Greenwich,
i.e., was shifted by 20° toward Japan, while its present location is not in
line with those tropical flora zones. In this scenario the flora of
Camtschatka, of the Amur region, and the Sachalin peninsula would be
located within the polar circle, those of Spitzbergen and Grinnell-Land
would even be north of the 60° latitude. The SUbtropical flora of Japan,
Alaska, the Mackenzie River, of Greenland and Iceland would be located
104 Neumayr: Die klimatischen Verhiiltnisse der Vorzeit [Climate Conditions of the
Prehistoric Age], op. cit., p. 27 of the separate print.
lOS Neumayr, op. cit., p. 30.
106 Nathorst: Zur fossi/en Flora Japans [The fossil flora of Japan]. Palaontologische
Abhandlungen, pub!. by Dames and Kayser, Volume IV., no. 3, 1888, p. 53,51 ff.
nico.stehr@zu.de
86 EDUARD BRUCKNER
between the 50° and 60° northern latitude, those of Switzerland, even more
tropical in nature, below the 36° northern latitude. Nathorst states that,
based on what we know today, we are led to assume that a shift in the
poles' location during the Tertiary Age seems most likely. Today, he is not
the only one to foster this view.
Independently from Nathorst, Neumayr came to similar results l07 and
provides paleontological facts in support of this theory not only for the
northern but also for the southern hemisphere. 108 Still, this does not solve all
problems, for in whatever position we put the pole within the circle of
tertiary fossil flora, the findings of tertiary forest trees will always be closer
to the pole than the northern tree line of today. Neumayr concludes that the
climate of the Tertiary was on the whole slightly warmer than today's but
to a much lesser degree than those countries benefiting from a shift of the
poles, such as Grinell-Land, Greenland, Spitzbergen, West- and Central
Europe, might suggest. 109
If we look back, we must admit freely that we know very little or almost
nothing about the climate conditions of the pre-diluvial geological past.
Only a few years ago the theory of a gradual cooling of the global climate
was the predominant one and it still has its followers; Heer's work on the
homogeneity of the climate and a gradual cooling during the Tertiary
Period was done in the 1870s and the early 1880s; and the most recent
work is that of Nathorst and Neumayr, who reject both those ideas. Even
today both views have supporters. Consequently v. Czerny fails to
recognise facts when he writes in 1881 that it is necessary now to explain
the causes for those changes which the climate has undergone since
prehistoric times, because these changes have long been a fact established
and proven by geologists. 110
Our knowledge about the climate during the diluvial age when the
glacial cover increased tremendously, is somewhat better; we will devote a
full chapter to this subject; therefore we can be brief at this point and
anticipate our findings recounted in that chapter.
It was once believed the remarkable expansion of the glaciers during the
diluvial period had local causes. Mountains were alleged to have been
much higher then and this played a major role. Charpentierlll mentioned
107 Neumayr: Erdgeschichte [Earth's History] II, p. 511-514.
108 Nathorst: op. cit., p. 55. Recently Schiaparelli did not fully exclude the possibility of a shift
of the earth axis, a possibility denied by many astrologers. Compare with Neumayr:
Erdgeschichte [Earth's History] II, p. 513.
109 Neumayr: Klimatische Verhiiltnisse der Vorzeit [Climate Conditions of Prehistoric Times],
p.38.
110 v. Czerny: Die Veriinderlichkeit des Klimas und ihre Ursachen [The Change of climate
and its Cause], Wien, Pest, Leipzig, 1881, p. 76.
III Charpentier in the Annales des mines, 1835.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 87
this idea at the beginning of his glacial studies and many others followed
his view. Nowadays however, after glacial research has discovered that the
ice age was a general phenomenon proportional in intensity to the size of
today's glaciers, this view must be considered refuted. There can be no
doubt that, during the ice age, global climate conditions were somewhat
different from today's.
One is generally inclined to assume that this difference refers mainly to
the amount of precipitation that was believed to have been much higher
during the ice age. That however is not correct; in our opinion the old view
from the year 1841 held by Agassiz and Charpentier, in which glaciers are
looked upon as thermometers, is not without warranty. We will show later
that even during the ice age the temperature must have been lower although
the difference to today's temperature was probably less than SoC globally.
Thus the ice age with its cool and wet climate occurs between the end of
the Tertiary Period whose climate conditions closely resembled today's,
and the next period. However there was not just one ice age but at least
two, and both were separated by a fairly dry inter-glacial age. Recurrent
climate variations are typical for the Diluvial Age and in strange contrast to
the prolonged gradual cooling of the Tertiary Period.
To this time we do not know what caused these climate variations
during the Diluvial Age nor do we know what caused the cooling process
during the Tertiary Period. So far, no convincing theory has been presented
which would be in complete accordance with facts. We will not revert to
speculations at this point as Croll, 1I2 Blytt,1I3 also Schmick114 have done,
though we do acknowledge these as initial efforts.
The climate of the ice age is different from today's: climate has changed
since the ice age. The change occurred when man was already present on
earth,1l5 and the question comes to mind immediately, whether man has
passed on reports of at least some of these events. It is the issue of climate
change in historical times that we are approaching.
Il2 Croll: Climate and Time (1875), as well as a number of recent publications.
I I3Blytt in Biologisches Centralblatt, Volume IV, p. 33 iT.
114 Schmick's numerous essays were listed in the Geographisches Jahrhuch, Volume V, Gotha
1874, p. 236.
115 Penck: Mensch und Eiszeit [Man and the Ice Age], Archive fUr Anthropologie, Volume
XV, 1884, Issue 3.
nico.stehr@zu.de
88 EDUARD BRUCKNER
4.1.2 Views and Opinions about Climate Change in
Historical Times
Many different hypotheses and theories about climate change have been
established in historical times, and have been defended more or less
successfully against never ending attacks. At one time general causes are
invoked, resulting in general climate changes without human interference; at
other times it is man's interference with nature and modifications to the flora
in particular, which are held responsible. In most cases it is a change in the
most important meteorological elements, such as temperature or rainfall, that
is claimed to have been observed. All other climate factors play only a minor
role in the literature on climate change. If occasionally the attempt is made
to emphasise, for instance, wind direction or wind velocity, it is almost
always done with the intention to gain better insight into the change of the
temperature or the rainfall.
Man's survival depends on rain, i.e., water, almost even more so than on
warm temperature conditions. Man can, to a certain degree, protect himself
against the cold by wearing appropriate clothing, but succumbs instantly to
arid conditions. Consequently, changes in rainfall have always been a
much-discussed issue in historical times.
Naturally, we are here mainly interested in those approaches trying to
verify the change in rainfall as it occurred since the end of the ice age.
Among the many works on this subject those of J. D. Whitney deserve
undoubtedly first rank. 116 According to his theory the ice age is a local
phenomenon, occurring for each single mountain formation, and a
necessary aspect of the general cooling process on earth. The latter, by
reducing the evaporation at the surface of the oceans more and more in the
course of time, is to be followed by a gradual drying out of the land masses.
This is supposed to be the gradual drying of the climate which, we are led
to conclude, is the cause for the observed gradual dissipation of the diluvial
glaciers. To establish this drying process from historical data was one of
Whitney's main goals. Of course, he is not the first to defend the idea of a
reduction in rainfall on earth caused by forces other than man's tampering.
Among Whitney's predecessors Theodor Fischer should be mentioned who
argued in several publications for the theory that the climate of the
Mediterranean countries had become drier since ancient times.1l7 But he
116 Whitney: Climate changes of Later Geological Times, Memoirs of the Museum of
Comparative Zoology at Harvard College. Cambridge, 1882.
117 Theobald Fischer: Uber Klimaanderungen an der A"quatorialgrenze der subtropischen
Regenzone [About climatic changes at the equatorial border of the rainy zone], Ausland,
1877, p. 891, published without name; Beitriige zur physischen Geographie der
Mittelmeerliinder, Leipzig, 1877; Studien iiber das Klima der Mittelmeerliinder [Studies
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 89
also, to some extent, blames deforestation, which has repeatedly changed
the face of the Mediterranean landscape completely during the course of
historical times. He feels, however, that the drying process and the
advancement of desert-like conditions towards the Mediterranean Sea is a
much too general phenomenon to be explained by such localised human
interventions, which moreover take place north of the 34° latitude. Rather,
he considers it to be an aspect of an, in general, increasingly drier climate
of the subtropical zones bordering the equator.
While Fischer limited his research to the surroundings of the
Mediterranean Sea, Whitney has gathered data on climate change in
historical times for almost all the countries of the globe. These data are to
be found in a special chapter together with the prehistoric geological
evidence of the climate's gradual drying trend since the ice age. Relevant
observations have been collected for the region of the Aralo-Caspian
Basin, for Persia, Central Asia, the Mediterranean Sea, the Sahara, Inner
Africa, and South America. He draws the conclusion: the climate has
become dryer all over the globe in historical times. This confirms the views
expressed for certain areas by many others,· apart from Fischer, namely by
Humboldt,118 Schmick,119 W. T. Blanford,I2O von Richthofen,121 O. Fraas,
Chavanne,122 lately by Jadrinzew,123 Venukof,124 W. Gotz,125 all of them
assume a development towards a drier climate unrelated to human
interference.
Varied are the facts from which conclusions for this drying process are
drawn. Only to a small extent is the evidence geological in nature insofar as
it is based on hydrographical changes which in all respects are beyond any
human interference, such as reduced water levels of lakes with no outlet;
therefore, the much larger diameter of the Lob-nor in the Tarym Basin
on the climate of the mediterranean countries], Petermann's Mittheilungen
Erganzungsheft No. 48. Gotha, 1879, p. 41 ff.; Zur Frage der Klimaiinderung im
sudlichen Mittelmeergebiet [About climatic changes in the Mediterranean area], etc.,
Petermann's Mittheilungen, 1883. p. I ff.
118 Humboldt for the Aralo-Caspian Basin in: L 'Asie centrale [Central Asia], Paris, 1843,
Volume II, p. 142.
119 Schmick: Die Aralo-Kaspische Niederung und ihre Behandung [The Aralo-Caspian Basin
and its Cultivation], Leipzig, 1874.
120 Blanford for Persia in Quarterly Journal of the Geological Society, Volume XXIX,
London 1873, p. 493.
121 v. Richthofen for the territory of Lob-Nor in: China, I, p. 110.
122 Chavanne: Die Sahara [The Sahara], Wien, 1879, p. 627.
123 ladrinzew about the disappearance of the West Siberian lakes, in Iswestija of the Imperial
Russian Geographical Society, XXII, No.2.
124 Venukofin Revue de geographie, X, Paris 1886, p. 81 fI.
125 W. Gotz: Die Verkehrswege im Dienste des Welthandels [The Traffic Routes Serving
World Trade], Stuttgart, 1888. p. 418, 506, 610, 669.
nico.stehr@zu.de
90 EDUARD BRUCKNER
about 4,000 years ago leads von Richthoven to the conclusion that aridicity
increased since that time. The disappearance of the Belt of Aibugir at the
Aral Sea as well as the allegedly continuous reduction of the Caspian Sea's
water level were once considered to be further examples of this process. To
a much larger part, however, the evidence relies on changes to plant life on
earth, the disappearance of some oases due to water-shortage for instance,
in short on appearances which are not entirely excluded from man's
influence and consequently are not complete or absolute proof of the
theory. As we will see below, they are to a large extent the same facts that
are used to prove climate change as a result of localised deforestation.
In fact, this theory of the climate's gradual drying has not escaped
opposition. The legitimacy of the assumption has been contested for almost
all of the regions that were claimed to be under the influence of a dryer
climate. It was pointed out that the fact that the shore lines of a lake with no
outlet are forced back, may well be attributed to a local siltation, and the
decline in the agricultural land of the Mediterranean regions to the
increasing indolence of their inhabitants. Just as determined as Fischer has
been in defending a climate change for the African regions of the
Mediterranean, Zittel 126 rejects this postulation, and Partsch sided with him
just recently, based on the extremely critical assessment of his research
material: The level of the Schott el Djerid in Tunisia has not changed since
pre-historic times, etc. 127 Tietze 128 voiced a similar opinion.
Very old and wide-spread is the opinion that forests have an important
impact on rainfall. And indeed, a priori, this seems quite likely. First of all
forests are natural barriers to wind-driven air masses, which are then, as
when encountering hills and mountains, forced to rise. No matter how light
this upward drift may be, in theory it will have to lead to more
condensation at its windward side. But the forest's influence is also felt in
the fact that the air above stays relatively humid. Forests slow down the
swift runoff of the rainwater and store the water in the ground, which is
then evaporated back into the air through the treetops. This process again
must bring about an increase in precipitation above the forest, the more so
as because of the strong friction between wind and forest surface and the
resulting delay in air flow the moist air tends to stagnate above the forest. If
126 Zittel: Betriige zur Geologie und Paliiontologie der Libyschen Wiiste [Contributions to the
Geology and Paleontology of the Lybian Desert], etc., Paliiontographica, Volume XXX, p.
42.
127 Partsch: Uber den Nachweis einer Klimaiinderung der Mittelmeerliinder [Evidence for a
Climate Change in Mediterranean Countries], Verhandlungen des VIII Deutschen
Geographentages [Discussions of the VIII German Geographical Convention]. Berlin,
1889, p. 123.
128 E. Tietze: Ueber Stepp en und Wiisten [About savannahs and deserts], Schriften des
Vereins zur Verbreitung naturwissenschaftlicher Kenntnisse, Wien, 1885, p. 160.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 91
forests enhance the amount and frequency of precipitation simply by being
there, deforestation as part of agricultural expansion everywhere, must
necessarily result in less rainfall and more frequent droughts. This view is
most poignantly expressed by the saying: Man walks the earth and desert
follows his steps! 129
Perhaps no other area on earth has been mentioned more often in
connection with the effects of deforestation on rainfall than the regions of
the Mediterranean Sea, and the increasingly drier climate since prehistoric
times that Fischer and Whitney interpreted as a general phenomenon, has
been attributed more frequently to man's localised destructive efforts of
turning woodland into arable land. And indeed, if we compare the
previously prosperous cultural life along the eastern shores of the
Mediterranean Sea with the Orientals' bare existence in those areas today,
we are struck by the tremendous cultural decline and are only too readily
inclined to see its causes in what saps our own energy upon arrival in the
orient: the scorching dry heat. To today's cultural leaders who live in the
cool, wet northern hemisphere it seems inconceivable that the blossoming
antique culture could have prospered under the present climatic conditions
of the orient: the climate must have grown warmer and dryer since
antiquity. Consequently, the clear-cutting of forests carried out in these
regions since ancient times offers a convenient explanation: man has
destroyed his own culture by destroying the forests and has devastated the
land, on which he is now left with a meagre existence. A comparison
between descriptions of the landscape as it used to be and as it is today
appears to confirm this situation. 130 Other authors voice their opinion in a
similar manner: Herschel, Arago, Klimtz, Lecoq, Clave, David Milne
Home, Mathieu, Wilson Flagg, G. vom Rath, Fautrat, Marsh, Simony,l3l
Denza,132 etc. Also Theobald Fischer believes that the drying of the
Mediterranean countries north of the 34th latitude is at least partially due to
deforestation. But it was often overlooked that in antique times the
129 Quoted by Simony: Schutz dem Walde [Protect the Forests!], Schriften des Vereins zur
Verbreitung naturwissenschaftlicher Kenntnisse, Wien. Volume LXX 1876177, Wien,
1877, p. 425.
130 A summary of the current literature on the influence of deforestation on the climate was
written by D. Milne Home in the Journal of the Scotish Meteorigcal Society, New Series
IV, 1870, p. 35 ff.; also LOffelholz-Colberg: Die Bedeutung und Wichtigkeit des Waldes
[The Impact and Importance of Forests]. Leipzig 1872; finally Whitney, op. cit. We only
refer to authors not named in any of the mentioned pUblications.
131 Simony: Schutz dem Walde! [Protect the Forest!], Schriften des Vereins zur Verbreitung
naturwissenschaftlicher Kentnisse, Wien, 1877, p. 451.
132 Denza: La meteorologia e la fisica terrestre at III, Congresso geografico internationale
diVenezia [Meteorology and geophysics at the III International Congress on Geography].
Roma, 1882, p. 16 fT. (Cited by Gunther.)
nico.stehr@zu.de
92 EDUARD BRUCKNER
descriptions of the subtropical environment were done by its inhabitants
while today's scientific efforts take mainly place in the temperate climate
zones of Europe. The inhabitant of a southern country had to see the same
phenomena with different eyes and painted them in different colours than
the inhabitant north of the Alps.
This increasing aridicity still continues today; Trottier 133 for example
substantiates this trend from 1838 on and especially since 1855 for Algeria
based on his observations of rainfall at Port d' Alger; and like Niel he
expects an improvement only through reforestation on a large scale.
According to Marmont,134 whose observations are repeated by N. Griigerl35
without quoting him, in Upper Egypt where rains used to be frequent just
100 years ago, it apparently does not rain anymore since the Arabs have
felled the trees in the mountains bordering the Nile valley to the east and
the west. In Grager's and also Anderlind'sl36 (1888) opinion, the opposite
result was achieved by extensive tree planting in the surroundings of Cairo
during this century, a view mentioned much earlier (1835) by Marmont.
Here, rainfall supposedly became more frequent whilst it was lacking
almost entirely before. Likewise, according to Murphy,l37 the area of the
Kidron creek near Jerusalem enjoys heavier rainfalls after a mulberry grove
was planted, etc.
It is not surprising that under such circumstances the issue of a link
between forests and climate has now and then been addressed by
governments. Lately, the Italian government has been paying special
attention to reforestation in Italy and its expected improvement of the
climate.138 Father Denza emphasises the goal of such replanting efforts with
these few words: It must be prevented that periods of heavy rainfall
alternate with droughts. 139
Indeed, observations about other areas seem to support these
conclusions regarding the forest's influence and the expectations tied to
reforestation.
It was often believed that the climate in Germany had improved since
the early ages, resulting in less cloud cover and precipitation from
increasing deforestation. And, in fact, a comparison of the rather gloomy
description of 'Germania', as for instance given by Tacitus, with the
133 Trottier cited by O. Niel: Geographie de I'Algerie [Geography of Algeria] 23 ed. T. I.,
l876,p.178.
134 Marmont referred to by Berghaus: Allgemeine Liinder- und Volkerkunde [General Geo-
and Ethnology], Volume II, Stuttgart 1837, p. 309.
135 N. Griiger: Sonnenschein und Regen [Sunshine and Rain]. Weimar, 1870, p. 153.
136 Anderlind in Meteorologische Zeitschriji, 1888, p. 154.
l37 J. J. Murphy: Are we drying up?, Nature X, 1876, p. 6.
138 David Milne Home, op. cit.; Gunther: Geophysik, Volume II, p. 290.
139 Denza, op. cit., p. 16 f( Cited by Gunther.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 93
Germany of today seemed to point to a climate change; it was not taken
into consideration though that this Roman's portrayal, naturally, had to be
subjectively distorted. But even in recent times multiple efforts have been
made to prove a link between changes in the climate and deforestation for
parts of Central Europe. Van Bebber expresses views along this line in his
publication on rainfall in Germany, 140 and so does Studni_ka for
Bohemia. 141 According to Wessely, in Hungary the climate of the steppe
has been gradually advancing since the lifetime of Maria Theresia. In his
opinion, resolute reforestation measures alone promise help in preventing
the impending drought. 142
In 1836 Riviere l43 advocated the theory of deforestation for parts of
southern France, namely the Vendee, the Provence, and particularly the
Department du Var, at the Academy in Paris; frost damage followed by the
clearing of olive tree plantations has presumably caused a considerable
reduction in rainfall and dried-up springs in the years from 1821 to 1822. A
similar situation exists in the former Poitou and the Department of the
lower Charente according to Fleuriau de Bellevue. l44 Actually, the question
of climate change due to destruction of forests has been raised in France
many times, i.e., in 1858 by Ladoucette, who pointed out before the French
Chamber of Deputies that the climate of the Departements Pyrenees
Orientales and the Herault had turned dryer and warmer after the
destruction of forests.145 Because of these reports the French legislature
took a serious look at the subject of reforestation. 146
As early as the l820s Kasthofer attributed the increasingly drought-like
ciimate conditions to the continued clear-cutting of forests in the alpine
areas of Switzerland and recommended reforestation. 147
In the Unites States deforestation plays an important role as well and is
seen as the cause for a reduction in rainfall, which is believed to have been
observed in the New England States and also in the Pacific States; 148 F. B.
140 van Bebber: Regenverhaltnisse Deutschlands [Rainfall in Germany], Miinchen, 1877,
p.119.
141 Studni_ka: Hyetographie von B6hmen [Hyetography of Behomeia], Archiv fUr
Landesdurchforschung von Bohmen. Volume VI. No.3. Prag, 1887.
142 Wessely in Simony: Schutz dem Walde! [Protect the Forest!], op. cit., p. 497.
143 Riviere: EfJets des defrichements, Comptes Rendus II, 1836,2, p. 358.
144 Fleuriau de Bellevue cited by Berghaus: Lander und V6lkerkunde [Geo- and Ethnology].
Stuttgart, 1837, Volume II, p. 30.
145 Hough: Report on Forestry, Washington, 1878, p. 395.
146 Marsh: The Earth as modified by human action, New York, 1877, p. 395.
147 Kasthofer, Bemerkungen auf einer Alpenreise [Notes from a Journey through the Alps].
Aarau, 1822. Annex: Klima des Alpengebirges [Climate of the Alps], p. 329.
148 Compare with Whitney's compilation: Climate Changes in later Geological Time,
Cambridge, 1882, p. 162 f.
nico.stehr@zu.de
94 EDUARD BRUCKNER
Houghl49 in his capacity as committee chairman of the American
Association for Advancement of Science demands decisive steps to extend
woodland in order to counteract the increasing drought.
Major droughts followed by famine characterised the sixties of this
century in Siberia. In general, one did not hesitate to link the drier climate
directly to the increased practice of deforestation. Van den Brinken even
blames the drought in the southern Russian steppe on the destruction of
woodland by nomads! These may be enough examples for the situation in
the temperate northern hemisphere.
Advancing deforestation was also thought to be responsible for the
purported dryer climate in some areas of the tropics.
Blanqui l50 considers the dry climate of the Isles of Cap Verdes to be the
result of deforestation, while claiming that St. Helena's rainfall increased
since the time of Napoleon I due to minor additions to the woodland.
Likewise, rainfalls reportedly have occurred more frequently on Ascension
after the British implemented reforestation on parts of the island. 151 The
island of Madeira lost all its forests through fire in the early 15th century
and as early as 1450 a reduction in rainfall was claimed to be noticeable. 152
Meldrum mentions in the late sixties that the inhabitants of the wetlands
on Mauritius cleared all forests with the purpose to gain dryer agricultural
land. However, clear-cutting went too far and resulted in terrible droughts
in the early sixties. Meldrum recommends reforestation to avoid their
reoccurrence. 153
In 1846 Gibson reported to the government l54 that according to the
native population India's climate in the areas surrounding Bombay and the
Nilgiri Hills has become drier since deforestation practices had increased.
According to Bidin the deforestation of the Coorg landscape in the western
parts of Ghats had a similar effect. 155 Just recently Blanford 156 established
the reverse effect for a region in central India's southern provinces: an
increase in rainfall coincides with an increase in woodland area
Reporting about his second expedition to New Mexico,Wheeler writes
that from year to year the climate becomes more desert-like because of
149 Hough, op. cit.
150 Blanqui is mentioned by Marsh (op. cit.) on page 184.
151 J. J. Murphy in Nature, Volume XV, p. 6.
152 Peschel: Neue Probleme der vergleichenden Erdkunde [New Issues in Comparative
Geography], Leipzig, 1870, p. 163.
153 Meldrum in Quarterly Journal of the Royal Meteorological Society, Volume IV, London,
1868, p. 187.
154 D. Milne Home, op. cit.
155 Bidin reports in Geographisches Jahrhuch, IV, p. 30.
156 Blandford: The Rainfall of India. Part II, p. 135 ff. and in Meteorologische Zeitschrift,
1888, p. 35.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 95
deforestation. In St. Cruz in the West Indies precipitation decreased to such
a degree due to deforestation that the island's wastelands and depopulation
figures are taking on sad proportions. 157 Sachs reports of similar situations
in parts of Venezuela's coastal region l58 and Hartt l59 does the same for
Brazil, where deforestation has already had an obvious impact on the
climate of Bahia.
Of the many examples of the situation in the tropics one may be
emphasised in particular since it seems to give compelling evidence of the
influence of forested land on rainfall and has been used as such many
times. We are talking about the observations made in the surroundings of
Lake Tacarigua or Valencia in Venezuela. Alexander von Humboldt visited
its shores in 1800 and verified that the lake's dimensions had decreased
since the founding of the city of Valencia, and in particular over the last 30
years of the previous century. Humboldt did not hesitate to attribute the
lake's descending water levels to the vast deforestation during the second
half of the 18th century. Twenty-two years later Boussingault visited the
lake and learnt that water levels had been rising again considerably for a
number of years; islands that had emerged in 1796 had disappeared again,
and large areas of acreage that used to be completely dry were in danger of
being flooded. Once again the forests are the answer. During the early years
of this century, bloody battles took place in Venezuela, particularly within
the immediate vicinity of the lake. Industry and agriculture were
deteriorating considerably and the lush growth of the tropical virgin forest
reclaimed the territory that man had taken away. As soon as the forest area
grew, the rainfall increased again and Lake Tacarigua began to rise. A most
compelling proof, indeed, of the forest's impact, one could not wish for a
more splendid one! 160
Analogous observations are available for the temperate climate of the
southern hemisphere. Milne Home l61 observed an area in Australia which
only twenty years ago had a precipitation of 37 inches; this amount is
now-Homes wrote in 1870-reduced to only 17 inches due to
deforestation. As a result, one of the Australian colonies took the initiative
and founded a special department for the preservation and planting of
forests. In fact, large-scale reforestation has been practised in certain areas
and Landsborough points out that Australia's climate is beginning to show
signs of an increase in precipitation and soon the continent will not only be
157 Simony, op. cit., p. 465 f.
158 Sachs is mentioned by Fritz in Petermann's Mittheilungen, 1880, p. 251 f.
159 Hartt: Geology and Physical Geography ofBrasil, Boston, 1870, p. 321 (quoted by
Whitney).
160 Boussingault: Influence des defrichements sur fa diminution des cours d'eau, Annales de
Chimie et de Physique. T LXIV, 1837, pp. 118-122.
161 D. Milne Home, op. cit.
nico.stehr@zu.de
96 EDUARD BRUCKNER
suitable for cattle farming but also for agriculture. 162 The climate in
Tasmania, on the other hand, as reported by Strzelecki, has become much
dryer thanks to the destruction of woodland. 163 According to S. Fritsch,
Wilson and Livingstone South Africa's Cape Colony is experiencing drier
climate conditions. Fritsch, who travelled in that area in the years 1863-66,
found traces of aridicity everywhere; 164 he sees its cause in increased
deforestation. B6hm and Bernsmann made the same observations in the
Herero Plains. 165
In looking at the views recounted above, whether they apply to parts of
the old or the new world, to temperate or subtropical zones, and in
consideration of the names of the scientists involved, there seems to be no
doubt: deforestation has played a major role in reducing rainfall in many
regions ofthe globe while reforestation has enhanced it.
This statement was substantially supported by research directed to
verify a decrease of the surface water in all countries with large
agricultures. The fact that spring waters are reduced or even completely
destroyed by deforestation has been known for some time. Becquere1
emphasised it,l66 as did Boussingault167 before him, the Swiss scientist
Marchand,168 and many others who have already been mentioned as
supporters of this theory about a link between deforestation and rainfall.
However, twice in this century the problem turned into a pressing topic
of the day: in the thirties and the early seventies. The former decade
produced an ample amount of literature about reduced water levels in the
rivers and streams of central Europe in the wake of the epochal research on
German rivers by Heinrich Berghaus,169 and particularly on the Rhein by P.
Merian. 170 From long-term observations over many, sometimes even 100
years, Berghaus determined that, overall, water levels of the rivers Elbe and
Oder had been falling continuously. He thought this reduction had to be
attributed to lower water volumes and related those to cultivation and
drainage of wetlands which used up more water due to evaporation. He
even expresses apprehensions that, if the trend continues, in 24 years, i.e.,
162 Landsborough: as reported in Nature, Volume XVI, 1877, p. 217.
163 Strzelecki: Van Diemens-Land, p. 192 f. (quoted by Milne Home).
164 More about Fritsch, Wilson and Livingstone in the essay of von Hann published in the
Zeitschrift fUr Meteorologie, 1869, p. 18 ff.
165 Bohm and Bernsmann in Petermann's Mittheilungen., 1878, p. 307 f.
166 Becquerel in Atlas meteorologique de I'Observatoire de Paris, 1867.
167 Boussingault: Rural Economy, p. 686. Cited by Whitney.
168 Marchand: Uber die Entwaldung der Gebirge [Deforestation of Mountains], Bern, 1849,
p. 29 ff.
169 Berghaus: Allgemeine Liinder- und VOlkerkunde [General Geo- and Ethnology], Volume
II, Stuttgart 1837, p. 300 fT. and p. 325.
170 Merian in Poggendorffs Annalen, LVII. 1842, p. 314 ff.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 97
around 1860, the water levels of the Elbe River will be too low for standard
boat traffic. That was in the year 1837; later he extended his research to the
Weser, Weichsel, and Memel and came to the same conclusions. 17l Both
times he identified the cause of the water level reductions in this way,
although putting some blame on deforestation in accordance with Pfeil's
theory172 and using a more cautious approach than most of his colleagues
who attributed the reductions entirely to the decline in rainfall due to
deforestation.
Those concerns were not just restricted to central Europe. A similar
water shortage occurred in Russia. The headwater regions of the Oka and
Don rivers, which were still rich in forests and waters during the first two
decades of this century, have become dry and barren according to a report
by the author A... written in 1842. 173 This situation is found almost
everywhere in Russia. In 1836 the low water levels of the Wolga raised
considerable concerns; they were attributed to massive deforestation and
threatened shipping operations. That is why Tsar Nicholas established a
commission to investigate the correlation between the reduction of
woodlands and the dropping water levels of the upper Wolga, which indeed
seems to have been confirmed. 174
In the early seventies G. Wex 175 published his well-known work about
the reduced amounts of water in springs, rivers, and streams. His research
material was by far more extensive than that of Berghaus but focused again
entirely on observations of water levels. Wex concluded that in cultivated
areas decreasing water levels result in a continuous decline in precipitation,
which in certain cases he even tried to measure. Based on observations of
the levels of the Rhein river near Basel during the years 1857-72 he arrives
at an annual drop of 1.97 em and based on this reduced amount of water at
171 Berghaus in his Annalen der Erd-, Volker- und Staatenkunde [Annals of Geography,
Enthnology and Political Sciences ], Series III, Volume IV 1838, p. 95.
172 Pfeil: Riihrt der niedrige Wasserstand der Fliisse etc. von der Verminderung der Wiilder
her? [Is the low water level of rivers, etc. the result of deforestation?], Berghaus' Annalen,
Series III. Volume IV, 1837, p. 289 ff.
173 A ... : About the causes of crop failures in Russia and Preventative Measures, Journal of
the Empirial Russian Ministerium for the Empire's Crown-Land and State Forests, 1842,
Part IV, p. 135 f., (in Russian). [Only the first character, "A", of the author's name is given
in the German original.]
174 See P. v. Koppen's report to the commission: in v. Baer and v. Helmersen: Beitriige zur
Kenntnis des russischen Reiches [Essays about the Russian Empire], Volume 4, p. III f.,
with a preamble by the editors which along with v. Koppen's report vehemently opposes
the commission's findings.
175 Wex: Ober die Wasserabnahme in de n Quellen, Fliissen und Strom en [Water Level
Reductions in Springs, Rivers and Streams], Zeitschrift des osterreichischen Ingenieur-
und Architektenvereins, 1874. Also Wex: II Abhandlung iiber die Wasserabnahme, etc. [II
Essay about water reduction, etc.], ibid., 1879.
nico.stehr@zu.de
98 EDUARD BRUCKNER
an annual decline in rainfall by 6.95 mm in the upper Rhein's catchment
area upstream from Basel. 176 His findings led Wex to the general
conclusion: a continuing decrease of the water levels of springs, rivers and
streams takes place in cultivated areas caused mainly by the increased
practice of deforestation and its resulting decrease in rainfall. M. W.
Schmidt came to the same results for the river Elbe.177
This proof had to raise some serious concerns. In 1873, in Wien, the
congress for agriculture and forestry discussed the problem in detail;178 and
when the Prussian house of representatives ordered a special commission to
examine a proposed law pertaining to the preservation and implementation
of forests for safeguarding, it pointed out that the steady decrease in the
water levels of Prussian rivers was one of the most serious consequences of
deforestation only to be rectified by reforestation programs. 179 It is worth
mentioning that at the same time or only a few years earlier the same
concerns were raised in Russia as well and governmental circles
reconsidered the issue of deforestation. 180
As often as the link between deforestation and rainfall is attested to, as
numerous are the voices opposing it, and it is difficult to decide who is
right.
It has only been a relatively short time ago that the section of
meteorology pertaining to forestry received a boost, in particular from
Ebermayer's work. 181 Not surprisingly, only a few reliable observations are
available regarding the influence of woodland on the climate in general and
on rainfall in particular. Those that are available seem hardly suitable to
support the theory of a decline in rainfall due to deforestation. Based on his
observations Ebermayer himself acknowledges that in plains or flatlands
with identical geographical characteristics, woodlands have very little
influence on the amount of rain; it increases however with higher elevations
above sea level.
Not all scientists were as cautious and prudent in drawing their
conclusions as Ebermayer was, and some achieved different results based,
in part, on verifiable fallacies. From their observations of rainfall over
forests and open country Fautrat and Sartiaux, for instance, attempted to
176 Wex: II Abhandlung, etc. [II Essay, etc.]'
177 M. W. Schmidt: Wasserstandsbeobachtungen an der Elbe im Konigreich [Water level
observations of the river Elbe in the Kingdom of Sachsen], Civilingenieur, Leipzig, 1878,
p.559.
178 Hough, op. cit., p. 292.
179 Schlichtung, in Deutsche Bauzeitung, 1875, p. 274.
180 Compare Journal of the Empirial Russian Ministry of the Empire's Crown Land and State
Forests. 1863, April, Report on Forestry, p. 167.
181 Ebermayer: Die physikalische Einwirkung des Waldes auf Luft und Boden, etc. [The
Physical Impacts Forests Have on Air and Soil, etc.], Berlin, 1873, p. 202.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 99
prove an increase in the amount of rain over the forest areas. 182 One
reviewer (Hann?) is however corred 83 in pointing out that the two
observation stations used in this comparison were operating under entirely
different conditions. The same can be said of analogous results by Mathieu
who over a period of six years carried out comparative observations in the
middle of a forest and in an area free of trees near Nancy.l84 His measuring
showed that the amount of rain that fell on the treed area was 6% higher
than the one that fell on the area without trees. However this result seems
quite shaky if one considers that the two stations, though being set up at the
same sea level, were positioned differently and 17 km apart. In general, all
attempts to solve the question under discussion through comparative
observations in a forested and a non-forested area can be contested on the
ground that they are prone to uncontrollable factors that may cause local
differences in rainfall, but have nothing to do with forests. This also applies
to A. W oeikof' s research-who has vigorously supported the forest
theory.18s
How numerous such local factors are, some of which can be attributed
to the placement of the measuring devices, and how much influence they
have on the difference in rainfall between two locations even within close
vicinity in a flat area with no elevations, has been proven through
observations initiated by Hellmann 186 and carried out at the test fields for
rain-gauging near Berlin. In 1886 and 1887 observations were made at 10
stations placed within a perimeter of just 40 square kilometres on a treeless
area with buildings and fields. Nonetheless the two-year averages of
precipitation differed by up to 16 percent. This shows that by using the
approach of comparative observations taken at the same time at different
stations, it is very difficult to prove the link between forest and rainfall.
Recently Blanford took steps to establish the effect of forested areas in
an approach we consider the only possible one. 187 We have already briefly
referred to his findings. In the southern part of India's central provinces
there is a large area which had been massively clear-cut in the past, but
where lately trees began to grow again. Deforestation ended and
reforestation began in 1875, and today all of this area is woodland again.
Blanford measured the average rainfall for the area in question at 14
182 L. Fautrat in Comptes Rendus, Volume 83, Paris, 1876, p. 514.
183 Zeitschrift flir Meteorologie, 1874, p. 384.
184 Compare A. Woeikoff: Einfluss der walder auf das Klima [Influence of Forests on the
Climate], Petermann's Mittheilungen, 1885, p. 81.
18S Woeikoff: Die Klimate der Erde [Climate on Earth], Volume I, Jena, 1887, p. 290 ff., also
several times in numerous publications of this highly reputable scientist.
186 Hellmann in Meteorologische Zeitschrift, 1887, p. [62] and in Wetter, 1888, p. 165 ff.
187 Blanford: The Rainfall of India, Part II, p. 135 ff., and in Meteorologische Zeitschrift,
1888, p. 35.
nico.stehr@zu.de
100 EDUARD BRUCKNER
stations from 1866-75, as well as from 1876-85, and determined a
substantial increase in precipitation. This increase was not noticed at seven
other stations close by yet outside the area of reforestation. He does not
hesitate to attribute this increase to the forest's impact, even more so since
it turned out to be a continuous process. Therefore, according to Blanford,
at least as far as tropical areas are concerned reforestation enhances rainfall
and deforestation, consequently, has to decrease it.
Gannet achieved entirely different results for the temperate North
American climate by using a similar method intended, again, to determine
quantitative changes in rainfall in areas that were undergoing extensive
gradual transformations of their vegetation. 188 First there are the prairies,
including Iowa, northern Missouri, southern Minnesota, most of Illinois,
and a small part of Indiana, an area of 5,000 square kilometres altogether,
where reforestation was done on the grandest scale. As it occurred on a
continuing basis, each series of observations if divided into two equal time
segments, should for one and the same location measure rainfall at a time
of little forestation in its first half, and at a larger extent of forestation in its
second half. Gannet discovered, however, that rainfall was 4% lower in the
second time period despite increasing reforestation. The opposite was
happening in Ohio and in the New England states where deforestation
continued. Still, Ohio showed a reduction in rainfall by only 0.5%, and the
New England states an increase of 7% based on the average measurements
from 12 stations done prior to 1860; the averages from 14 stations after
1860 showed no change at all. Gannet concludes from these data that
deforestation and reforestation has no significant impact on the amount of
precipitation.
On the one hand, based on their respective experimental research
Ebermayer and Gannet refuse to accept the theory that deforestation brings
about a substantial change in the amount of rain, at least not in temperate
climate zones, while Blandford seems to have proof of the opposite for the
tropics. On the other hand, the discussion of long-term meteorological
observations has by no means resulted in a confirmation of the acclaimed
reduction in rainfall either. SchottI 89 and Draper,l90 for instance, point out
that there is no evidence of such a reduction in the eastern part of the
United States over the last 60 years. lamieson 191 shows in 1859 that rainfall
has not changed since the middle of the last century in Great Britain.
188 Gannet's essay is discussed in detail in Das Wetter, 1888, pp. 97-105.
189 Schott, in Smithsonian Contributions to Knowledge, Volume XXIV, Washington, 1885,
p.288.
190 Draper, in Meteorologische Zeitschrift, 1874, p. 239.
191 Jamieson mentions Wex's research in his report to the Wiener Akademie: Sitzungsberichte
der Wiener Akademie, II Abtheilung 1874, p. 642. A lecture was also published in Kiimtz'
Repertoriumfor Meteorologie [Repertory for Meteorology].
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 101
Burton l92 and Schweinfurth l93 shared his view with regard to
Egypt-contrary to Anderlind et al. The commissions of most of the
academies in Europe to which Wex had submitted his report about the
reduction in water levels and rainfall in cultivated regions for appraisal
could not from meteorological observations in their own countries support
this view-despite increased deforestation. This was the response of the
academies in Paris,194 Petersburg,195 Wien,l96 and also of the commission of
the Austrian Society of Engineers and Architects. 197 However, very few
accurate meteorological observations date back further than the middle of
the past century so that they are at best predicative for the duration of one
century. Efforts have been made to find out more about the changes in
rainfall in earlier time periods by comparing meteorological observations
that have been carried out without instruments. By comparing
meteorological notes from the diary of Tycho de Brahe with today's
meteorological data, La Cour infers that cloud and rain patterns have not
changed over the past 300 years at the coast of the Sound. 198 And R. Wolf
concluded from a lecture held by J. Gessner in 1747 that during the past
140 years the average amount of precipitation in northern Switzerland did
neither increase nor decrease. l99 These samples are, however, too small to
serve as conclusive evidence. As far as exact observations are available, we
are able to draw the following conclusion from the research we introduced
above: Despite the fact that deforestation has taken place to this day in ever
growing proportions, according to meteorological records rainfall has not
decreased in those regions. Any reductions in the water levels of rivers
should be linked to other factors. In what way this may be done has for
instance been pointed out by A. Marie Davy.2°O He again attributes the
falling of water levels in France to the expansion of cultivated land but
relates it not to less precipitation but to an increase in evaporation which is
highest over cultivated land and higher than over woodland.
192 Burton: The Gold Mines of Midian, London, 1878, p. 26.
193 Schweinfurth, in Biidecker: Agypten [Egypt], Volume I, 1878, p. 79.
194 The Report of the Paris Academy mentioned by Grebenau in Deutsche Bau-Zeitung" 1876,
p.426.
195 Bulletin de I'acad. des sc. de Petersbourg. T. XXI, 1876, pp. 293-302.
196 Sitzungsbericht der Wiener Akademie, mathematisch-naturwissenschaftliche Classe,
Volume LXIX, Part 11,1874, p. 642.
197 Report in the journal of this society, 1881, p. 86.
198 La Cour: Tycho Brahe 's Meteorologische Dagbog holdt paa Kransborg for aarene
1582-1597 [Tycho Brahe's meteorological diary from Kransborg for the years
1582-1597], Kopenhagen, 1876.
199 R. Wolf, cited by Giinther: Geophysik II, p. 294.
200 Marie Davy, in Meteorologsiche Zeitschrift, 1874, p. 145 ff.
nico.stehr@zu.de
102 EDUARD BRUCKNER
This disproves the meteorological conclusions Wex had drawn from the
claimed reduction in water levels in accordance with the deforestation
theory. Even this reduction, the basis of Wex's inquiry, was declared
without proof. In a very critical essay written in the thirties even PfeiFo l
cautioned against blindly accepting the observations by Berghaus and
others about sinking water levels as the symptom of an actual reduction in
flowing bodies of waters. K. E. v. Baer and Helmersen,202 and also
Berghaus 203 agreed with him and declared the forests not responsible. Later
technical aspects were occasionally added to the discussion initiated by
Wex's inquiry. For instance, 1. Schlichting/04 Sasse/os and others point out
that only then will water levels accurately reflect the quantity of the
flowing water if the cross-section of the river and speed of flow have not
changed. That, however, is an assumption that will never apply to bedload
rivers. Such changes are actually produced by correcting the flow of the
body of water, which in this century has affected our rivers enormously.
Grebenau 206 for instance, interpreted the general reduction in water levels,
on the average by 1 cm annually, as the result of lower river beds, as an act
of erosion. FesseF07 expresses the same opinion on this subject.
Others again claimed that the lower averages could have been caused by
a change in the regimen of the rivers due to increased deforestation and
drainage of the river catchments. Deforestation has completely changed the
way rainwater drains: now rain runs off much faster than before when the
20t Pfeil: Rilhrt der niedrige Wasserstand der Fli1sse und insbesondere derjenige der Elbe und
der Oder, welchen man in neuerer Zeit bemerkt, von der Verminderung der Wiilder her?
[Are the low water levels of the rivers and the Elbe and Oder in particular, which have
lately been noticed, caused by deforestation?] Pfeil's kritische Bliitter fUr Forst und
Jagdwirthschaft, Volume XI, 1837, issue 2, p. 62 ff.; reprinted in Berghaus's Annalen der
Erd-, Volker und Staatenkunde [Annals of Geography, Enthnology and Political Sciences],
Series III, Volume IV, 1837, p. 289.
202 Baer and Helmersen: Beitriige zur Kenntnis des russischen Reiches [Essays on a better
knowledge o/the Russian Empire], Volume 4, Petersburg, 1841. Preamble to the Report of
P. v. Koppen to the Commission about his inquiry into a possible link between the
decrease in woodland and the decrease in the water of the upper Wolga river.
203 Berghaus: Allgemeine Liinder- und Volkerkunde [General Geo- and Ethnology], Volume
II, Stuttgart 1837, p. 310.
204 Schlichting, in Deutsche Bauzeitung, 1875, p. 273. Also compare with the reports of
academies and societies mentioned above.
205 Sasse, in Deutsche Bauzeitung, 1873, pp. 259,268.
206 Grebenau: Resultate der Pegelbeobachtungen an den elsass-lothringischen Fli1ssen Rhein
und Model von 1807-1872 [Results of Water Level Observations at the Alsatian Rivers
Rhine and Moselfrom 1807-72], Strassburg, 1874 (III Heft der statistischen Mittheilungen
iiber Elsass-Lothringen) Also: Flusssenkungen und die amit zusammenhiingenden
Erscheinungen [Lowered river levels and related phenomena]. Lecture in Deutsche
Bauzeitung, 1876, p. 425.
207 Fessel, Deutsche Bauzeitung, 1873, p. 329.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 lO3
forest's soil absorbed it like a sponge and released it gradually into the
streams. As a result, the off-seasonal fluctuations of water levels have
become more frequent and stronger, the floodwaters more numerous and
higher, the low waters more frequent and lower. Such a process must
necessarily result in lower annual water levels. For every river the water
level corresponding to the annual mean stream flow is higher than the
average water level; for, the stream flow is not linearly proportional to the
depth but proportionally to a higher power. Only in rivers without any
annual fluctuations are both water levels, the one corresponding to the
average rate of flow and the arithmetic average of all water levels per year,
the same. However, the greater the difference between the levels of low-
water and high-water, the more the average annual water level will fall
below that corresponding to the average stream flow. Therefore, as a result
of deforestation the annual water levels will be lower while the stream flow
is unchanged. The hydro-technical committee of the Austrian Society of
Engineers and Architects 208 commented on the subject along these lines, as
did von Helmersen and Wild,209 as well as Whitney,2IO Markham, Hann,2l1 v.
Wagner,212 and others. Deforestation changes the regimen of bodies of
water and consequently their water levels, but does not affect rainfall or
water volume-that is the gist of it.
But even this relatively small role which forests play in this process has
been successfully and expertly disputed by rejecting the fundamental fact
from which Berghaus and Wex proceeded, namely the continuous decrease
in the water levels of streams in cultivated areas. Schlichting pointed out
that parts of W ex's own data do not even show a continuous lowering of
water levels if one determines the averages in a different way than Wex
did.213 As far as a reduction undeniably did occur, it might partly be
explained by the fact that ice blockages have become less frequent and
smaller. 214 Hagen mentioned that a reduction of water levels is only found
208 Report in Zeitschrift des 6sterreichischen Ingenieur- und Architektenvereins, 1875, p. 157
ff.; conclusion.
209 v. Helmersen and Wild's commentary on Wex's study. Bull. de l'acad. des sc. de St.
Peterbourg, 1876, p. 293 ff.
210 Whitney op. cit. , p. 179 ff.
211 Hann: Thatsachen und Bemerkungen iiber einige schiidliche Folgen der ZerstOrung des
natiirlichen Pjlanzenkleides der Erdoberjliiche. [Facts and comments on some damaging
results of the destruction of plant life on earth], Meteorologische Zeitschrift, 1869, p. 18
ff. Also Markham is mentioned here.
212 v. Wagner: Hydrologische Untersuchungen an der Weser, Elbe, dem Rhein und mehreren
kleineren Fliissen. [Hydrological Research of the Weser, the Elbe, the Rhine and a number
of smaller rivers], Braunschweig, 1881, p. 24.
213 Schlichting, in Franzius and Sonne: Handbook of Engineering, Volume IV: "Der
Wasserbau" Leipzig, 1883, p. 73 of the second edition.
214 Schlichting, in Deutsche Bauzeitung, 1875, p. 144.
nico.stehr@zu.de
104 EDUARD BRUCKNER
in some of Prussia's rivers, and then again in others it is not,215 and by
comparing observations done at several gauging stations at the Elbe,
Pralle216 as well as Schlichting217 showed that the lowering of the water
level is a local occurrence while in other spots along the same river levels
are rising. No doubt this points to local changes to the riverbed and not to
less water in the river. Graeve came to the same conclusions--only earlier.
At the same time he emphatically opposes Grebenau' s theory that rivers in
general are running at lower levels. He invalidates the theory simply by
pointing out that, if it was correct, almost all rivers, not long ago, would
have been running well above the average level of the older cities and
communities along their paths. 218 Recently Honsell expressed his opposition
to the idea of a general reduction of water levels in rivers. Thus, neither
Wex's nor Grebenau's theory is necessary, since what both wanted to
explain, the general decrease of water levels in the rivers of cultivated
areas, does not exist and the observed fluctuations of water levels are
discontinuous, irregular events related to local conditions.
Let us review the many hypotheses mentioned! Deforestation is the
cause of a dryer climate everywhere, it reduces the water levels in our
springs, creeks and rivers-this is the opinion of one group; no trace of a
drier climate and no trace of lower water levels-argue the others. Two
opinions that are mutually exclusive and yet their advocates are first-rate
scientists. At present we cannot decide between the two. Only one thing is
evident: We are still very much in the dark as to the link between forests
and rainfall.
This becomes even more evident when we learn that deforestation not
only reduces rainfall and river water, but supposedly also increases both in
some blessed spots on this earth. I am not referring to the ancient story by
Theophrast according to which, as Seneca reports,219 the river Haemus had
plenty of water after deforestation-but to more recent authors who have
written on this subject. Strangely enough voices were raised favouring an
overall increase in river levels, and again deforestation had to be the culprit.
As Schmid writes in 1858, it was widely believed that due to the practices
215 Hagen: Dber die Verminderung der Wasserstiinde in den preuj3ischen Str6men [Reduced
water levels in rivers in Prussia], Abhand1ungen der Koniglichen Akademie der
Wissenschaften in Berlin, 1880.
216 Pralle: Wasserstandsverhiiltnisse in der Oder [The Situation of the Odra's Water Levels],
Zeitschrift fUr Bauwesen, 1882, p. 188.
217 Schlichting: Elbestromschau 1869 und 1873, [Inspection of the river Elbe 1869 and 1873],
Deutsche Bauzeitung, 1975, p. 274.
218 Graeve: Dber den Wasserreichthum und die Senung der Flusse in Culturliindern [Water
volume and its decline in the rivers of cultivated], Deutsche Bauzeitung, 1877, pp. 261,
271 ff.
219 Seneca: Quaestiones naturales Ill, 11.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 105
of deforestation and amelioration in Poland the river Weichsel now (1858)
carried more water than before. 220 However, he argues against this theory
and relates the acknowledged rise of the water levels to major ice blockages
caused by dyking. His argumentation is similar and yet reversed to
Schlichting's, who linked the observed decrease in water levels to a smaller
amount of ice blockages.
We have mentioned before, that the increasing dryness of Australia's
climate had in the past been related to massive deforestation. But according
to the latest research carried out in New South Wales and other regions of
Australia this assumption is entirely incorrect, as mentioned by R. Abbay,221
M. E. Abbott222 and R. von Lendenfeld. 223 The opposite has been observed
in many regions where since the late sixties it was deforestation that
increased the area's water resources. For example, Lake George and Lake
Bathurst, two lakes in New South Wales with no outlet, rose considerably
since the fifties. According to Abbay, obviously rain water was able to
reach the lakes much faster than in the past without evaporating after
forests had been cut down. Many dry or sparsely running creeks have
turned into permanently flowing murmuring brooks after the valleys'
forests were cut down. Consequently, Abbott and Lendenfeld recommend
to further extend the practice of deforestation in order to provide dry
regions with more water veins. "This way Australia would be able to draw
more and more amounts of rain water and at least in part conserve them".
One explanation for this phenomenon, so incompatible with findings from
the 'old world', may be the fact that Australian trees absorb larger amounts
of water which dries out the forests' soils more quickly.224 But it
nevertheless seems strange that the same deforestation of Australian
woodlands which in previous years was generally held responsible for the
increasing dryness of the climate, should now attract rain and bring water to
the soil. One is inclined to assume that, as so often, the truth is caught
somewhere in the middle and that deforestation has been as innocent of
causing less rainfall and lower water levels in the past as it is today. In that
case, increase as well as decrease of rainfall occurs irrespective of
deforestation for completely different reasons. In any case, this unusual link
between deforestation and climate in Australia, which at times reverses,
220 Schmid: Nachrichten uber die Strome des preuj3ischen Staates [News about Prussia's
Rivers], III; Weichselstrom, Zeitschrift fUr Bauwesen, VIII, 1858, p. 158 f.
221 Abbay in Nature, XIV, p. 47 f.
222 Abbott in the Publications of the Royal Geographical Society ofN. S. Wales, referred to in
American Meteorological Journal, Volume IV, 1887, Oct., p. 247.
223 R. v. Lendenfeld: Der Einfluj3 der Entwaldung auf das Klima Australiens [The Impact of
Deforestation on Australia's Climate], Petermann's Mittheilungen, 1888, p. 41 ff.
224 von Lendenfeld, op. cit., p. 43.
nico.stehr@zu.de
106 EDUARD BRUCKNER
certainly proves the point we made earlier-that we know nothing about
the influence of forested areas on rainfall.
The situation seems not to be any different regarding another human
interference with the climate conditions on this earth. I am referring to the
improvement of the climate by expanding the cultivation of originally dry
regions and areas with very little vegetation. This view dates back to
ancient times; Theophrast reports that the water sources and lakes around
the city of Arcadia on Crete dried up when agriculture was abandoned. But
after the city had been destroyed and the land was once again used for
agriculture, water, too, came back. 225 Lately, this hypothesis was applied on
a grand scale to the dry regions in the Western United States of America.
In this case, the facts from which such far reaching conclusions
regarding climate change are drawn, are not meteorological in nature;
rather they have to do with the expansion of agriculture into areas which 30
years ago were considered totally unsuited for any kind of cultivation. In
1856 the eastern border of the infertile "Great American Desert" ran
between the 96th and 97th meridian through the States of Dakota,
Nebraska, Kansas, the Indian Territories and Texas. Gradually settlers
moved beyond this line and step by step agriculture was pushed further
west. Today, it has exceeded the lOOth meridian and partly reached the
102nd. The official 1885 census for Kansas puts the population in the
sector between the 97° and 100° latitude of this state at more than half a
million, who settled there during the past 20 years. It is the fundamental
belief of the settlers that they have changed the climate and increased
rainfall by cultivating the dry land and planting wheat and corn. They
attribute the transformation of the land and its climate to their own
energy.226 Similar cases are reported from Montana and Dakota. 227
The same happened around the Great Salt Lake. Here it was observed
that from 1860 on, rivers were rising; one was able to drain their water for
the irrigation of fields, and agriculture took possession of regions that only
a short time ago were considered infertile. 228 The Great Salt Lake rose by
more than 3 m and its area increased from 4,532 sq. km to 5,609 sq. km.
The rising water level of the lake coincided with the expansion of
settlements around the Lake and the increase in rainfall and water volume
225 Ideler: Ober die angebliche Veriinderung des Klimas [About the alleged changes of
climate]. Berghaus: Annalen der Erd-, Volker und Staatenkunde [Annals of Geography,
Enthnology and Political Sciences], Vulume V, 1832, p. 425.
226 Compare with the lecture of Heyer about N. R. Hilton: Report of the Kansas State Board of
Agriculture, for the quarter ending March 31, Topeha, 1888, in Wetter, 1889, p. 223 £ The
original was not available to me.
227 American Meteorological Journal, Volume IV, 1887, Oct., p. 242.
228 See also Gilbert, in Powell: Report on the Lands of the Arid Region of the United States,
2nd Edition, Washington, 1879, p. 57 cont'd.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 107
in streams and lakes is directly attributed to the expansion of cultivated
land recovered from the desert. In 1869 Cyrus Thomas commented on the
noticeable improvement of the climate over the past 8 years;229 he is
convinced that as the population increases so will the rainfall. In 1878
Hough voices a similar opinion. He states that the diligent Mormons have
good reason to expect further increases in rainfall if cultivation continues;
with more trees planted, the air will become more humid until finally
precipitation will be sufficient. 230 Such statements reflect common opinion.
Whether rainfall actually increased or not has evolved into a big
controversy.
Gilbert, the unremitting explorer of the Great Basin, comments on the
cause of water increase in the Great Salt Lake region in the critical manner
peculiarly his own. He leaves open whether it is the result of a climate
change or of the settlers' changes to the land's drainage structure. 23i If the
former should be the case, he rules out anything but a general climate
change that came about without any human effort and is comparable to
geological climate changes. He considers the changes in the water
resources to be constant over longer periods of time and not just the sign of
a short-term oscillation of the climate around a mean.
The view that rainfall has increased on the wide plains of the far West is
lately being shared by Morrow, Snow, Ch. F. Adams and Greely, while
Dorsey232 and also H. A. Hazen 233 were unable to see the increase from
looking at observations of the rainfall, and S.R. Thompson234 declares the
entire issue unfit for judgement at this point because it has not been fully
investigated. Three recently published inquiries try to approach the problem
in a more exacting manner by using meteorological data.
It is. evident that a decision can be reached immediately by comparing
two rain charts composed of data collected from different time spans. Mark
W. Harrington has taken this approach by comparing two rain charts of the
United States, the one created by Blodget on the basis of older observations
prior to 1856 and the other by Ch. Denison based on observations by the
Signal-Service during the years 1870 to 1883. He discovered that the
isohyetes of the prairie areas between the 45° and 30° northern latitudes
229 Gilbert, op. cit., p. 71.
230 Hough, op. cit., p. 92.
231 Gilbert, op. cit., pp. 67-77.
232 Harrington refers briefly to Adams, Snow, Morrow, Greely, and Dorsey in American Met.
Journal, Vol. IV, p. 369; and Curtis ibid., Volume V, p. 69 ff.
233 H. A. Hazen: Variations of Rainfall West of the MiSSissippi River. Signal Service Notes. N.
VII. Washington, 1883.
234 Thompson, in American Meteorological Journal, Volume I, p. 59.
nico.stehr@zu.de
108 EDUARD BRUCKNER
have undergone a general shift to the west, which corresponds to an
increase in rainfall. 235
Gannet achieved the exact opposite result. Applying the method
described earlier (p. p.) he concludes that in the region between Missouri
and the Rocky Mountain range precipitation did not increase and
consequently the cultivation of the land had no influence on rainfalp36 The
latest research by G. E. Curtis led to the same negative result. 237
Some time ago Whitney argued strongly against the hypothesis that
increasing cultivation of the land enhances rainfall;238 to him it appears
unthinkable that around the Great Salt Lake the cultivation of a mere 11400
of the total area (equal to 1112 of the lake's area) should have been able to
increase rainfall and raise the lake's level to such an extent as seems to be
the case. Moreover neighbouring regions show increased water levels as
well, even though they experienced deforestation but no cultivation.
Powell, again, rejects this view in his article published at the end of 1888.239
We are observing the same conflict of opinions here as in the issue of
the link between forests and precipitation. First it is claimed that the climate
in the interior of America became more humid thanks to the practice of
cultivation; then again it is stated that agriculture is not responsible; and
finally an increase in rainfall cannot be established at all.
Let us summarise the results of our brief historical excursion by
outlining in a few words the present status of the issue concerning a change
of precipitation in historical times.
An increase in rainfall in historical times is nowadays claimed for a few
limited regions only and in most cases related to some form of human
intervention.
On the other hand, the number of researchers who opted for a decline in
precipitation is quite impressive. However, the cause of this decline is a
matter of completely different opinions. Some are willing to blame
everything on increased deforestation without knowing much about the
effect of woodlands on precipitation. Others, led by Whitney, reject the
idea that deforestation could have such a major effect on rainfall and
assume instead that the global climate is in the process of drying up and the
many single occurrences are to be considered as nothing more than
symptoms of this development.
235 Harrington, in American Meteorological Journal, Volume IV, p. 309 ff. Compare with my
article in Meteorologische Zeitschrift, 1888, p. 43.
236 Gannet, op. cit., p. 103.
237 Curtis: The Trans-Mississippi Rainfall Problem Restated, American Meteorological
Journal, Volume V (June 1888), p. 66 ff.
238 Whitney, op. cit., p. 179.
239 Powell, in Proceedings ofthe Royal Geographical Society, London 1888, p. 793.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 109
The strongest opposition to all these researchers comes from those
scientists who deny any change in rainfall in historical times. It is
noteworthy that the majority of meteorologists, whose observations do not
date back very far, belongs to this group of opponents.
A consensus between these fundamentally different views seems
impossible and we cannot help but ask ourselves: how could this confusion
of opinions come about?
The number of researchers who favour the theory that the temperature
has changed in the course of historical times is anything but small. The
majority of these do not rely on temperature readings but on symptoms
evident in the flora and fauna on earth, or on changes in hydrographical
phenomena which have been discovered by comparing the historical data
from different time spans.
The fact that today's population of Greenland is rather small compared
to the 15th century, and above all that Greenland's east coast is said to have
been inhabited as late as the end of the middle ages while today those
regions are covered with ice and surrounded by an almost impenetrable belt
of pack-ice, has frequently been used as evidence for a trend towards a
cooler climate in historical times. It was talked about a shift of the
isotherms to the south toward the North Atlantic Ocean. This view is not
new and has been repeated many times up to this date. It was strongly
reinforced by the term "Greenland" which indeed is not a very suitable
name for the Greenland of today. Arago mentions this view in his essay
about temperatures on earth/40 as does Bernhard Studer. 241 Even more
recent authors such as Czerny242 and lately Michelier43 have tried to draw
major conclusions from this supposition that Greenland's climate was so
much milder in the past. Whitney244 again insists that a climate change took
place in Greenland, yet uses a more critical approach than his predecessors.
Nonetheless, today, all these conclusions have essentially lost their basis
since Conrad Maurer established that there never have been any Norman
settlements on Greenland's east coast and that those small Norman hamlets
on the West Coast built in the 10th century and resembling Eskimo
settlements did not succumb to a climate which had become inclement, but
240 Arago: Oeuvres completes. T. VIII. Paris 1858, p. 243.
241 B. Studer: Lehrbuch der physikalischen Geographie II. [Textbook ofphysical geography).
Bern, Chur, Leipzig 1847, p. 306.
242 Czerny: Veriinderlichkeit des Klimas [Variability of the Climate], Wien, Pest, Leipzig
1881, p. 5.
243 Michelier: Etude sur les variations des glaciers des Pyrenees [A study about the vaiations
fo the Pyrenean glaciers]. Annales du Bureau Central Meteorologique de France, 1885.
Part I, pp. 207-234.
244 Whitney, op. cit., see p. 239.
nico.stehr@zu.de
110 EDUARD BRUCKNER
to an invasion by Eskimos against which they were-abandoned by the
motherland--quite defenceless. 245
It has often been claimed that the European climate shows some signs of
a general cooling trend. This idea seemed to harmonise perfectly with
Greenland's climate change. The decline of cultural life in Iceland, for
instance, is linked to such a temperature change,z46 where, in addition, the
tree line is said to have retreated indicating a deterioration of the climate.
Signs for this cooling trend are apparently to be found all over Northern
Europe and Asia.247 Tree lines are retreating southward on the Shetland
Islands, in Iceland and Scotland according to Czerny,248 in Lapland
according to Whitney,249 in Siberia according to v. Middendorf250 and F.
Schmidt. 251 Similar observations were made for the Alps by a number of
researchers, such as Kasthofer 52 and TschudF53 for Switzerland, whose
findings are unconditionally confirmed by Theobald,254 M uret,255
Leresche,256 and Coaz,257 as well as Whitney for the entire Alps.258
A retreating tree vegetation is almost always interpreted as a sign of
climate change. Ideler however maintains as early as 1837259 that a shift of
the polar tree line to the south may well be the result of human intervention
because the largest and best trees are continuously cut down leaving the
smaller trees exposed to the wind and making their survival much harder
than it used to be. Coaz 26O draws similar conclusions. He attributes the fact
that mountain tree lines are descending to lower elevations, to the damage
caused by alpine farming and its herds of cattle, but not to changes in
temperatures. L. Dufour agrees and rejects the conclusion that this fact was
the necessary outcome of a climate change.
245 Compare with article by Kirchhoff about Czemy, Leopoldina 1881, p. 176.
246 Whitney, op. cit., pp. 239 ff. and 236.
247 Ideler, in Berghaus' Annalen, 1832, Volume V, p. 421.
248 Czemy, op. cit., p. 49.
249 Whitney, op. cit., p. 236.
250 v. Middendorf: Sibirische Reise [Siberian Journey], Volume IV I, St. Petersburg, 1867,
p.612.
251 F. Schmidt: Resultate der zur Aufsuchung eines Mammuthcadavers etc. ausgesandten
Exepedition [Results of an Expedition sent to Explore the Remains of a Mammoth, etc.]
Bulletin of the St. Petersburg Academy, Volume VIII, 1872, p. 26.
252 Kasthover: Bemerkungen auf einer Alpenreise, etc. [Notes from a Journey through the
Alps, etc.] Aarau, 1822.
253 Tschudi: Die Alpen [The Alps] 1859, p. 305.
254 Theobald, in the 1868 lahrbuch des Schweizer Alpenclubs.
255 Muret, in Dufour, Bulletin Soc. Vaud. des sc. X., p. 373.
256 Leresche in Dufour as above, p. 378.
257 Coaz: Letter to Dufour-in Dufour, op. cit., p. 379.
258 Whitney, op. cit., p. 236 f.
259 Ideler, op. cit., p. 420 f.
260 Coaz, op. cit., p. 375.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 111
At the end of the last century Hamilton advocates this change towards a
cooler climate for Great Britain and Ireland relying on the judgement of
farming experts. 261 Glaisher showed however, that London's temperature
has risen considerably over the past 100 years. But it turned out that
Glaisher's conclusion applies to London only, the temperature increase
being a result of the ongoing expansion of the city, which in time
completely enclosed the meteorological observation site. 262 Finally, based
on a series of long-term temperature readings Buchan came to the
conclusion that the temperature in Scotland had not changed at all since
readings began. 263
One country which has been the focus of especially thorough and varied
investigations regarding the issue of climate change, particularly in
historical times, is France. The identified changes are in part contradicting.
According to Picot France's climate has become considerably warmer since
early history as a consequence of the gradual deforestation in this
country.264 The same view was later expressed in Government Chambers by
Ladoucette for the southern part of France?65 Arago,266 on the other hand,
draws the opposite conclusion from the fact that in the past there used to be
wine growing much further to the north; he claims that in the course of the
past centuries France's summers became cooler and winters warmer. And
once again increasing deforestation is blamed! In agreement with this
hypothesis and based on phenological observations Bourlof67 tried to
establish in 1870 that the climate of the Alsace had deteriorated
considerably since the 13th century. Of course, Arago does not suggest that
this climate change took place in all of France but in most of it. He refers to
the development in other countries where deforestation had the same effect
on temperatures. Apart from the fact that Arago's allegation is inconsistent
with any observations made so far about the link between forested areas
and temperatures, even the preposition for his findings, that is the shift to
the south of once northerly wine-growing regions, is nowadays according
to Ideler and Ch. Martins 268 attributed to other than climatic impacts. In
France, just as in Germany and England where in previous centuries
vineyards were cultivated in areas of higher latitudes than today, their
261 Hamilton, in Transactions of the Irish Academy, Volume II, 1788; cited by GUnther:
Geophysik II, p. 294 (was not available to me).
262 See Whitney, op. cit., p. 228.
263 Buchan: Climate a/Scotland, Athenaeum, 1876, p. 329.
264 Picot, cited by Ideler, op. cit., p. 425.
265 Ladoucette cited by Hough: Report on Forestry, Washington, 1878, p. 293.
266 Arago: Oeuvres completes. T. VIII. Paris 1858, p. 230 ff.
267 Bourlot in Bull. d l'Ass. sc. France, 1870,23 January.
268 Ideler, op. cit., p. 449; Ch. Martins: Le climat de la France a-t-if change? Annuaire
meteorologique de France pour 1850, p. 121.
nico.stehr@zu.de
112 EDUARD BRUCKNER
relocation to the south is nowadays merely seen as a symptom of a more
refined taste and a better transportation system allowing good quality wine
to reach the markets at lower prices and over larger distances.
While Picot, Arago and others, in one way or the other, fully supported
the theory of a change in the French climate, others denied it ever having
occurreq. In a critical assessment Charles Martins, as we have mentioned
earlier, rejected Arago's view as being unwarranted. But he also disagreed
with the opposite hypothesis and pointed out that Roman records, if used to
suggest a climatic improvement, must be read with reservations and can
only be looked upon in relative terms-just as the excellent Ideler did 13
years earlier. By examining meteorological observations Schouw69 and
Dove refuted the occurrence of a continuous change in the temperatures of
Germany, Denmark, and Scandinavia while Zimmermann270 later claimed
again that the climate in and around Hamburg had cooled off. For Geneve
Gautier 71 and Plantamour72 showed that since the middle of the last
century not even the trace of a change in temperature could be
substantiated.
Arag0 273 claimed to have established a pattern of climate change for
Tuscany in Italy with cooler summers and warmer winters, which he again
attributed to increasing deforestation. But Whitney274 rightly points out how
unreliable the meteorological data collected in the 17th century are, which
Arago used. Temperature changes in some other regions of the SUbtropical
hemisphere have been reported as well. It is primarily the fact that in
Afghanistan the northern growth line of the date palm tree retreated further
south, which was interpreted in this way, as for instance by Bellew275 and
Whitney.276 Theobald Fischer prefers to link this event to human negligence
in cultivation and plantation efforts.277
As in France, two very different rivaling hypotheses about climate
change are prevalent in North America. At the end of the last century
Larochefoucauld-Liancourt278 supported the assumption that in Canada
269 Schouw: Skildring af Vejrligets Tilstand i Danmark [Description of the weather in
Denmark, Kjobenhavn, 1826. Cited and referred to by Ideler, op. cit., p. 428 ff.
270 Zimmennann, in Poggendorff's Annalen 1856, Volume 98, p. 323.
271 Gautier, in Bib!. Univ. de Geneve, 1843, Janvier, T. XLIII, p. 158.
272 Plantamour: Le climat de Geneve [The climate of Geneve] , Geneve.
273 Arago, op. cit., p. 227.
274 Whitney, op. cit., p. 234.
275 Bellew: From the Indus to the TigriS, London, 1874, p. 239.
276 Whitney op. cit., p. 233.
277 Fischer: Die Dattelpalme, [The date palme] Erganzungsheft, No. 64 of Petermann's
Mittheilungen, Gotha 1881, p. 50.
278 Larochefoucault-Liancourt: Voyage dans les Etat-Unis de l'Amerique septentrionale, [A
Voyage tothe septentrional United States], Volume II, p. 207.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 113
summers were getting hotter and winters milder. Much earlier Peter Kalm
expressed a similar conviction for parts of the United States, as mentioned
by Volney (1803).279 In contrast, according to S. Williams 280 and
Williamson,281 who were publishing at roughly the same time as the authors
just mentioned, and according to general public opinion summers have
allegedly become considerably cooler in the New England states, in
Virginia (Jefferson), and in Louisiana (Thomassy)282 and on the whole the
climate has become more temperate. The development is similar for
Newhaven as Loomis and Newton283 discovered on the basis of
meteorological observations. All these temperature changes are almost
unanimously attributed to excessive deforestation or, in the case of
Newhaven, to local reforestation. However, these conclusions are
conflicting with findings by many researchers who strictly dispute any
change in temperature such as Humboldt, Noah Webster, and Forry,284 and
recently Schott285 backed by a large sample of meteorological data.
Draper's286 objection is based on the observation that the length of time
during which the Hudson River is covered with ice in winter remained
unchanged.
Apart from all this evidence of a temperature change in historical times
there are, beside the ones included above, additional data supporting the
idea that temperatures remained constant. A comparison between
descriptions of the cultural life and the plant life at antique historical sites
and what life is like today led Arag0 287 to dispute any change in temperature
for Palastine, Syria, Egypt, and Greece. By using the same method E.
Bioe88 concluded the same for China; according to him the temperature has
not changed in this country in 3300 years. Ideler's and recently Partsch's
findings about the climate in Mediterranean countries are along the same
lines. 289
279 Volney: Tableau du climat et du sol des Etats-Unis d'Amerique, Paris, 1803, T. I. p. 288.
280 Williams and Jefferson, as referred to by ldeler, see p. 422.
281 Williamson mentioned by Draper in Zeitschrift fUr Meteorologie. 1874. p. 240.
282 Thomassy, mentioned by Dufour, op. cit., p. 364.
283 E. Loomis and H.A. Newton: On the Mean Temperature and on the Fluctuations of
Temperature at Newhaven, Transactions of the Connecticut Academy of Arts and
Sciences, Volume I, p. 194. (Cited by Whitney.)
284 Humboldt, etc., mentioned by Draper, op. cit.
285 Schott: Tables, Distribution and Variations of Atmospheric Temperature in the United
States, Smithsonian Contributions to Knowledge, Volume XXI, p. 311.
286 Draper in an article in zeitschrift der osterreichischen Gesellschaft fUr Meteorologie,
Volume IX, 1874, p. 239 ff.
287 Arago, op.cit., p. 222 ff.
288 E. Biot: La temperature ancienne de la Chine, 1841 (quoted by Dufour).
289 Partsch: Verhandlungen des VIII. Deutschen Geographentages zu Berlin, Berlin, 1889.
nico.stehr@zu.de
114 EDUARD BRUCKNER
Most of all we have the excellent research by L. Dufour about the
question of climate change in Switzerland. 29O His critical stance in the
discussion about why the Alps' tree lines have dropped to lower altitudes
has already been outlined above. He also shows that all the information
concerning the past growing of olive trees or grapes in regions where today
they do no longer thrive, is in part misunderstood, in part unreliable and in
contrast with other information, and can, finally, in part be explained by
human interference unrelated to the climate. Generally, all evidence for a
climate change in Switzerland that is based on observed changes in the
vegetation, is not conclusive because these changes can be explained
differently. However, it is nevertheless quite remarkable that these
phenomena, as far as they were interpreted as symptoms of climate change,
should all point to a deterioration of the climate, to a decrease in
temperature. The registered dates for the harvesting of grapes in
Switzerland point to the same result. In the sixteenth and early seventeenth
century grapes were harvested much earlier in the year than they are today
and particularly than they were in the eighteenth century. As interesting as
this fact may be, Dufour does not fail to recognise that even the timing of
this harvest may depend on reasons which have nothing to do with the
climate. He, therefore, leaves the question whether the climate has changed
or not, wide open. And in fact A. Angot has recently shown that this delay
in the grape harvests of past centuries did not occur in the neighbouring
Swiss Jura and in the Departement Cote d'Or.291
As a result the question whether temperatures changed or remained
constant in historical times is today as far from being solved as the question
of changes in precipitation. Some say the climate is getting warmer, then
again others say it is getting colder. The explanation for the alleged
temperature change was sought in completely different causes. Nowadays,
of course, no one would relate a cooler climate to increased deforestation,
as Arago did. But deforestation is still held responsible for a warming
trend. On the whole however, one is definitely more inclined to attribute
climate change to general causes. Two hypotheses oppose each other in this
regard. Schmick compiles the data favouring a warmer climate and relates
this temperature change, which according to him is typical of the northern
hemisphere, to the earlier onset of the equinox. 292 In contrast, Whitney talks
290 Dufour: Variation du climat [Climate variations], Bull. Soc. Vaudoise des Sc. Nat., X,
pp. 359-436.
291 A. Angot: Etude sur les vendanges en France, Annales du Bureau Central Meteorologique
de France, 1883, Part I, p. [B.] 83.
292 Schmick: Die Aralo-kaspische Niederung im Lichte der Lehre von den siicularen
Schwankungen des Seespiegels und der Wiirmezonen. [The Aralo-Caspian Basin in Light
of the Theory of Secular Variations of the Lake's Level and the Temperature Zones],
Leipzig, 1874.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 115
about a general global cooling process which manifests itself in historical
times all over this earth and is nothing else than the cooling trend which
dates back to the beginning of the Tertiary and continued through the
earliest geological periods.293 As in the past, these views are still challenged
today by a number of scientists, among them the majority of the
meteorologists, who categorically insist: Temperatures remained
unchanged in historical times.
In view of the legions of hypotheses about a change in temperature or
rainfall the number of attempts made to verify changes in any other aspect
of the climate is rather small. Nevertheless, assumptions regarding changes
in the wind pattern of certain regions for example have been made,
defended and attacked. According to S. Williams and Jefferson, in New
England the frequency of westerly winds has decreased while easterly
winds occur more frequently.294 And again this altered state is attributed to
deforestation. According to Simony the Bora winds have become stronger
in such clear-cut areas with karst topography.295 Lespiault attempted to
verify a major climate change in France which he ascribes to the growing
force of those storms attacking the West Coast of France. 296 Blavier
suggests the exact opposite: a decreased wind activity and an increasingly
calmer atmosphere combined with more fog which he attributes to a
potential change of direction of the Gulf Stream along the French coast. 297
GruB identifies certain areas in Europe where wind directions have changed
in the course of this century.298 In Miinchen, for instance, since 1865 north
winds are presumed to blow less often than in previous years, easterly and
westerly winds, however, increased; in the city of Leipzig westerly and
northerly winds appeared to have increased slightly; in Berlin easterly and
north-westerly winds increased as well, while in Lund it is the northerly
and easterly winds which occur less often in this century as opposed to the
last.
We have arrived at the end of our tour. We have walked through a
veritable labyrinth without the benefit of Ariadne's clew. Again and again
we encountered the same insoluble and irreconcilable contradictions. It
seems almost like a psychological puzzle, that for one and the same country
serious scientists have at every step insisted on climate changes which are
293 Whitney, op. cit.
294 S. Williams and Jefferson, reviewed by Ideler, op. cit., p. 426.
295 F. Simony: Schutz dem Walde! [Protect the Forest!], Schriften des Vereins zur
Verbreitung naturwissenschaftlicher Kenntnisse, Wien, Volume XVII, 1876-77, 1877, p.
456.
296 Lespiault, cited by GUnther, in Geophysik, II, p. 289.
297 Blavier: Changement du climat sur les cotes de la Vendee etc. L'Astronomie (de
Flammarion), 1883, p. 106 ff. Cited by GUnther. (The original was not available.)
298 GruB, in Wetter, 1888, p. 137, and in Meteorologische Zeitschrift, 1888, p. [57] No. (155).
nico.stehr@zu.de
116 EDUARD BRUCKNER
mutually exclusive. Equally puzzling is the fact that over and over again the
forest is singled out as the scapegoat for a variety of frequently
contradictory changes. Something quite striking emerges from all this: the
lack of any progress towards a solution. Opinions are still as directly and
diametrically opposed today as they were forty years ago. If the climate
would measure up to all these assumptions which have been uttered over
the past years and decades, it would have to change back and forth from
one direction to the other. When we look back on this welter of different
hypotheses, we have to admit that even today we are still far from a definite
answer to the question whether the climate changed in historical times or
remained constant. And even today we still have to adhere to L. Dufour's
statement of some twenty years ago: "The question of a climate change in
historical times is still wide open and the claim by the majority of
meteorologists that climate does not change has found neither more nor less
proof than the claim to the opposite.299 " If the climate is indeed undergoing
a continuous change, it is undoubtedly occurring at a very slow pace, which
would explain why even today we have no definite knowledge of its
specific aspects.
4.1.3 Meteorological Cycles
Parallel to the described research into climate change in historical times,
efforts were going on in the past centuries to find meteorological cycles, or
secular climate variations in predetermined time periods, motivated partly by
hopes to gain a safe forecasting method for the future. These approaches
went into two different directions and neither their methods nor their
outcome had anything in common. At one time this turned into an
unsystematic hunt for periodical data that led to tabulations of a variety of
diverse cycles. Then again it was attempted to prove the connection between
periodical phases of the sun and similar phases for meteorological elements
on earth. This created the wealth of literature about the influence of sunspots
and their II-year periods on the meteorological conditions on earth.
On earth a change in weather is governed by the cyclical change of
annual seasons; it seemed therefore rather obvious to conclude that there
might be similar cycles of much longer duration, e.g., resembling an annual
period of a higher rank, which show a regular pattern of weather changes
from one year to the next. The search for periodic weather patterns has in
fact always been a thankless job in meteorology. The method of
299 L. Dufour, op. cit., p. 420.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 117
establishing and verifying such cycles has most often been totally
inadequate. A series of brief reports published by Symons' Monthly
Meteorological Magazine in April and May 1886 may serve as an example.
Brumham, for instance, claims to have identified a 40-year period of cold
and warm winters, but in such a way that several of these periods are
concurrent. Hot and dry summers return after 425 years; G. T. Gwilliam
establishes a return of every 17 years. In a second publication Brumham, on
the other hand, identifies cycles of 8, 10, 12, 19, 29, 39, 40, 68, and 136
years for the occurrence of hot summers and forecasts one for 1886,
because that particular year would fall into all those cycles with the
exception of the 19-year period. The subsequent cool summer of 1886, it
would seem, taught him a lesson. Data on which such an approach is based
are less than adequate.
Even some distinguished meteorologists have tried to find a periodicity
in weather patterns. But more often than not their arguments do not bear up
against serious criticism. The highly regarded director of the
meteorological network of New South Wales, H.C. Russel, for instance,
advocates a 19-year weather period;3°O that, however, is not supposed to
mean that every 19 years rainfalls or temperatures reach a maximum, but
rather that each year shows the same characteristics as the corresponding
one of 19 years back; unfortunately his evidence is inconclusive. Benj.
Gould believed to have discovered an 18-year periodic pattern of the wind
velocity for Buenos Aires.
The search for a periodic pattern in the occurrence of cold winters is a
frequent endeavour. We have already mentioned one such case. But as
early as 1741 the Petersburg scientist Krafft stated categorically that
extreme winters tend to occur every 30 to 34 years;301 he tried to back up
his findings with less than complete statistical data about severe winters in
historical times. Endeavours like this have been attempted until recently,
i.e., in 1876 by Chavanne who, based on data about the ice cover of the
polar sea, identified periodic reoccurrences at a multiple of three, i.e., every
3,6,9,12,15, 18,21, etc., years. 302
All such attempts have one thing in common: they cling too much to the
single year and see the weather as a process developing with an almost
300 H. C. Russel: History of Floods in the River Darling, Journal and Proc. R. Society New-
South-Wales for 1886, Sydney, 1887, p. 156 ff.
301 Krafft: Ausfohrliche und umstiindliche Beschreibung des im Januar-Monat 1740 in St.
Petersburg errichteten Eispalastes, etc. [Detailed and lengthy description of the 'Ice
Palace' constructed in St. Petersburg in January 1740], St. Petersburg, Academy of
Science, 1741, in Russian, pp. 23-29.
302 Chavanne: Die Eisverhiiltnisse im arktischen Polarmeere und ihre periodischen
Veriinderungen [The ice conditions in the arctic polar sea and their periodic variations],
Petermann's Mittheilungen, 1876, p. 254 ff.
nico.stehr@zu.de
118 EDUARD BROCKNER
mathematical precision. Deviations, of course, do occur, which are declared
random occurrences; yet it is not checked why these deviations are in fact
lower in number than they ought to be, if the arrangement of years was left
to chance. As soon as the random rule is applied, the alleged regular pattern
collapses.
A more scientific approach of verifying a periodic pattern is based on
multi-year averages. But even then fallacies are easily committed if the
number of deviations is not carefully examined in the above manner. This
is still evident in a note by A. Duponchel published in the Comptes Rendus
in 1888;303 he determined the 12-year averages of the annual temperature of
Paris starting with the year 1804, e.g., from 1804 to 1815, 1816 to 1827,
etc., and was surprised to find out that two successive time periods each,
"in general" deviate from the multi-year average under opposite signs;
indeed, based on this 24-year pattern he even feels inclined to predict a cold
winter for 1896/97 and a very mild one for 1908/09!
In examining observational data from large meteorological networks
scientists occasionally claimed to have discovered certain cycles here and
there. Wild advocated that St. Petersburg's temperature changed in 23-year
cycles. Later he also found a cycle of approximately 40 years for the
rainfall measured at Russian stations. 304
A multi-year period of cold winters was found by E. Renou/ 05 but not in
the same sense as did the authors mentioned above. This cycle implies that
a particularly severe winter occurs every 41 years surrounded by a 20-year
period with numerous winters which are not quite as severe but still cold,
while the remaining 20-year period shows relatively few hard winters. In
essence, he assumes that periods with a great number of cold winters
alternate with periods with a lot of warm winters. Similarly, Koppen
suggests a 45-year period of severe winters for the past two centuries (18th
and 19th) and a l30-year period for earlier centuries. 306
From observations in Praha, Milano, Wien and Miinchen Hornstein
believed to have identified a 70-year period regarding air pressure. This
period coincides with a period of sunspots. A 22-year period of temperature
changes, as well as a 7-year one in particular, was identified by Schott for
303 Duponchel, in Paris Comptes Rendus, 1888, 2 e semestre, p. 427.
304 Wild, Temperaturverhiiltnisse des Russischen Reiches [Temperature conditions o/the
Russian Empire], Supplementary Volume to Report. f. Met. St. Petersburg, 1881, p. 279;
also Wild, Regenverhiiltnisse des Russischen Reiches [Rain Conditions 0/ the Russian
Empire], Supplementary Volume to Report. f. Met. St. Petersburg, 1887, p. 80.
305 Renou: Periodicite des grands hivers, Annuaire de la Societe Meteorologique de France
1861,p.19ff.
306 Koppen, in Zeitschrift ftir Meteorologie, 1881, p. 183 ff.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 119
the United States. However, the length of this period varies a little as the
case may be. 307
Of all those meteorological cycles, which oc((ur in a more or less
established pattern, none has been treated with the same versatile and
scientific interest as the incidental II-year period of meteorological
phenomena that was assumed to have been the result of an II-year period
of sunspot activity.308
Soon after Schwabe and R. Wolf had identified the periodical re-
occurrence of sunspots, the search for an II-year temperature cycle started.
Previously, Gautier had attempted to establish a 10-year temperature cycle
for a number of stations in connection with the 10-year period of sun spot
activity discovered by Schwabe in 1843. For the first time in 1853 Fritsch
established an II-year temperature cycle for seven European stations.
According to both scientists, years of minimal sunspot activity are marked
by very warm temperatures. Subsequently, in numerous publications
several other researchers have tried to investigate this link further, i.e.,
Zimmermann, Piazzi, Smith, Stone, Hill, Hahn, who came up with positive
results, while Celoria for Milano, Hann for Wien and Schott for the United
States failed to discover a connection between sunspots and temperature;
Baxendell, Weilenmann, Blanford, Roscoe and B. Stewart claimed to have
found a correlation, in which, however, periods are swapped. According to
their theory the temperature development is not inverse to the sun spot
activity but runs parallel to it.
Among all related research, Koppen's inquiries into multi-year weather
cycles are the most outstanding. Even today, Koppen's findings have not
been surpassed. According to these results, during the time span from
1816-1860 sunspot activity and temperatures correspond globally like
mirror images as Gautier and Fritsch had asserted, but prior to 1816 or after
1860 this accord is sporadic.
Just as numerous are the attempts to establish an ll-year cycle for the
rainfall. In 1872 Meldrum was the first to refer to a periodical pattern of
307 Schott: Tables, etc. of the Atmospheric Temperature in the United States, Smithsonian
Contributions, Volume XXI, No. 277, p. 314.
308 Hahn gave a summary of the position on this question: Uber die Beziehungen der
Sonnenjlecken zu meteorologischen Erscheinungen. [About the Link Between Sunspots and
Meteorological Phenomena], Leipzig, 1877. H. Fritz: Die Beziehungen der Sonnenflecken
zu den magnetischen und meteorologischen Erscheinungen der Erde, [The Link Between
Sunspots and Magnetic and Meteorological Phenomena on Earth], Haarlem, 1878; v.
Czemy: Die Veriinderlichkeit des Klimas etc [The Variability of the Climate etc.], Wien,
Pest, Leipzig, 1881, p. 9 ff.; the latest and most detailed summary was given by v. Bebber
in his Handbuch der ausubenden Witterungskunde [Manual ofApplied Meteorology],
Volume I, Stuttgart, 1885, pp. 199-257. A well-documented bibliography is included. We
can therefore refer to names without quotations.
nico.stehr@zu.de
120 EDUARD BRUCKNER
cyclones in the Indian Ocean south of the equator whose maxima and
minima correspond with those of the sunspot activity. As a sequence he
later tried to establish that the rainfall is slightly heavier during times of
maximum sunspot activity than during times of a minimum. Lockyer,
R. Wolf, Symons, Hunter, Brocklesby and H. Fritz agreed with Meldrum's
theory while Celoria, B. Stewart, Strachey and Whipple, and in part also
Jelinek were unable to verify the correlation from the data they had worked
with. In general, however, it is quite likely that rainfall increases with an
increase in sunspot activity. A most significant result was achieved by Hill
and E. Douglas Archibald who independently established that the summer
and the winter rains in India show an entirely different correlation. The
summer rains reach maximum amounts at the time of a sunspot minimum,
the winter rains at a sunspot maximum in accordance with the continental
and dry regions on earth.
It has also been attempted to relate the remaining meteorological
phenomena to sunspots. Hornstein, Forssmann, Hahn, S. A. Hill, H. F.
Blanford; F. Chambers, Douglas Archibald and J. Allen Broun have found
a periodic pattern of barometric pressure corresponding with the II-year
cycle of sunspots. The correlation appears to have been established for
Southeast Asia implying that the higher barometric readings are in
accordance with the maxima-the lower with the minima. However, where
the process of compensation takes place, which is required due to the static
nature of the total air mass, has not yet been examined. The frequency of
cyclones, it appears, increases with that of the sunspots, complying with the
correlation established by Meldrum and defended by A. Poey. The same
increases occur in wind velocity, according to Riihlmann, while the
influence on wind direction has not yet been verified. The same applies to
clouds, thunderstorm activity and hailstorms.
Some went even further and tried to establish the influence of an
assumed II-year weather cycle on hydrographical and even on economical
aspects. ReiB, for instance, insisted on a link between periodic floods and
sunspots. An II-year cycle had been claimed earlier, though partly with
little success, by Dawson for the oscillations of the American Great Lakes,
and by Fritz for the fluctuations of rivers in general and even the sizes of
the European glaciers. Hunter talks about an II-year cycle of famines in
India, Javons about one pertaining to commercial crises, etc.
In looking at the current status of the sunspot issue it cannot be denied
that there are indeed close links between the different meteorological
phenomena and periods of sunspot activity. However, the physical aspect
of the argumentation for this correlation is in part in a sad state, as, for
instance, regarding rainfall and even in respect to temperature. For, even
the simple question whether a sun radiates more heat with or without spots,
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 121
has not been solved. It remains equally unexplained why a meteorological
phenomenon appears to concur with the phase of sunspots in one location
and not in others, and why such a concurrence continues for some time in
one place and then again disappears. In short, the issue still remains in the
dark and unsolved. This can be said of the II-year cycle of sunspots and
much more so of the longer 55-year cycle; its influence on our atmosphere
is at this point quite hypothetical apart from the occurrence of Northern
Lights and of magnetic phenomena which are not part of this study.
It was the search for a clearly defined cycle which governed the theories
thus described and which, in the case of the sunspot cycle at least, led to
results. All other theories of cycles have in part no scientific basis
whatsoever. Entirely different are the views we would like to address now
and which in their majority were produced during the past years.
The whole question of climate change in historical times entered a new
phase when it was no longer the continuous change in one direction, be it of
the rainfall or the temperature, which was sought or the short weather
periods of a certain duration, but when meteorological data were instead
examined with regard to secular ups and downs of the climate over long
periods of time.
This was initiated by the unusual variations in the size of glaciers.
Meteorological conditions had to be the sole cause because the very
existence of glaciers is directly linked to certain climatic conditions; their
change must necessarily cause a change in the size of glaciers; there could
be no doubt about that. For a long time glaciers had therefore been looked
upon as a kind of thermometer or measuring device of weather conditions
in general. The more striking was the fact that this obvious relationship
between glacier size and weather variations could not be verified by
meteorological observations until in 1858 v. Sonklar309 spoke the relieving
words and regrettably no one listened. In his excellent report he provided
proof of the parallel development of glacial variations and variations in
temperature and rainfall over a time span of two centuries. For the more
recent time period he used the meteorological observations of stations in
Milano and HohenpeiBenberg; for the earlier years he collected as many
data as he could obtain about weather conditions of individual seasons as
far as they pertained to the territory of the Alps and their surrounding area.
Since general weather accounts as well as precise observations were
available to him regarding the last 100 years, he tried to quantify such
general terms as "cold", "very cold", etc. with the help of a clever, yet
309 v. Sonklar: Ueber den Zusammenhang der Gletscherschwankungen mit den
meteorologischen Verhiiltnissen, [About the relationship between glacial variations and
meteorological conditions], Sitzungsberichte der Wiener Akademie, Volume 32, 1858, pp.
169-206. Including a diagram displaying curves.
nico.stehr@zu.de
122 EDUARD BRUCKNER
somewhat controversial method. He obtained proportional figures which
increase and decrease according to changes in the two factors mainly
responsible for the formation of glaciers: cold temperatures and moisture.
von Sonklar recognised quite clearly that it is not the weather pattern of a
single year that determines glacial variations but rather the consecutive
weather patterns of many years. Consequently, he deemed it necessary to
adjust the meteorological data in order to be able to demonstrate the
parallel development of glacier variations and weather changes. He was
able to achieve this by ignoring individual years and their quite irregular
and random weather changes and by concentrating instead on their five-
and ten-year averages. In neutral ising the influence of individual years in
this manner, the path was set which generally had to lead to success.
von Sonklar's results are clear and exact. Based on both, the general
weather accounts as well as the measurements from observational
instruments, he determined that in the Alps the glacial expansions around
1770, from 1810 to 1820, and in the forties of the 19th century coincided
with periods of wet and cool weather, whereas the reduction of the glacial
cover by the end of the 18th century and in the twenties, thirties, and fifties
of the 19th century coincided with periods of dry and warm weather.
But v. Sonklar's work went unnoticed and the question about the causes
of glacial changes continued to be looked upon as unanswered until the
publication of Forel's research in 1881, which at least prepared the ground
for a possible solution. Independently of v. Sonklar, but using a similar
method of neutralising the swings of the annual weather patterns Forel
achieved the same results. From his observations in Geneve he was able to
verify that those periods when glaciers expand are indeed immediately
preceded by periods of low summer temperatures and heavier precipitation,
whereas glaciers diminish in size when the weather is quite warm and
dry.3lO
Two years later the same route was taken by Eduard Richter in his
accomplished case study of the Obersulzbach glacier. 3Il He made use of the
rainfall measurements of Klagenfurth for his purposes by determining the
5-year averages in an attempt to show the general tendency of this climate
factor. His conclusion was that the rainy periods of 1842 to 1852 caused the
glacial expansion in the fifties, while the dry periods of 1852 to 1872 were
the cause of the unusually large latest reduction. He finds it remarkable
though that the significant increase in rainfall from 1872 to 1878 has not
yet shown any effect on the glacial cover.
310 Forel: Variations periodiques des glaCiers, Archives de sc. phys. et nat. Geneve, 1881,3.
Per. T. VI, p. 22 and p. 451.
3Il E. Richter, in Zeitschrift des deutschen und Osten'eichischen Alpenvereins, 1883, p. 75 ff.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 123
These findings were confirmed in a more general aspect by C. Lang,
who, by adhering strictly to the methods chosen by Forel, extended the
study to all of the surrounding regions of the Alps using rainfall
measurements from nine and temperature measurements from five stations
primarily located at the foot of the Alps and in the Alps' foreland. 3I2 Here
again, the link between rainfall, temperature and glacial variations showed
a definite parallel tendency.
Given the importance of the research by Forel, Richter and Lang for the
issue which we will explore later on, we may be permitted to tabulate their
results below. Von Sonklar's findings could not be included because he
does not separate between temperature and rainfall. Maximum, respectively
minimum precipitation occurs in the following years:
Maximum PreciEitation Minimum PreciEitation
Geneve (Fore!)313 1842/57 1878/80 1835/41 1858177
Klagenfurth (Richter) 1842/52 1872/78 1852172
Milano (Lang) 1810/14 1840/49 1880/84 1825/29 1865/74
Praha 1815/19 1845/49 1820/24
Wien 1875/79
Milnchen 1850/54 1880/84 1870/74
Hohenpbg. 1805/09 1835/44 1820129
Reichenhall 1845/49 1855/59
Stuttgart 1845/54 1880/84 1860/64
Chioggia 1800/09
Geneve {ForeQ314 1826/35 1856/75 1836/55
Maximum TemEerature Minimum TemEerature
Milano (Lang) 1790/94 1820/29 1845/49 1810/19 1835/44
Stuttgart 1825/34 1860/64 1835/44
Regensburg 1775/79 1795/99 1825/34 1815/19 1835/44
Milnchen 1830/34 1865/69 1840/64
Hohenpbg. 1790/94 1820/24 1810/19 1835/39
312 C. Lang: Der siiculare verlauf der Witterung als Ursache der Gletscherschwankunegn in
den Alpen [The secular weather pattern as a cause of glacial variations in the Alps],
Zeitschrift ftiT Meteorologie, 1885, p. 443 ff.
313 It must be kept in mind that the annual figures determined by Forel are not directly
comparable with Richter's and Lang's findings. The figures for the first period of heavy
rainfalls from 1842-57 must be interpreted in such a way that the 10-year averages of
1833-42, 34-43, ... ,48-57 were higher than the multi-year average. The correct way of
putting it would therefore be: heavy rainfall from 1837 to 1852, if the first 10-year average
of 1833-42 is connected to the middle year 1837, the last of 1848-57 to the middle year
1852.
314 Only according to summer temperatures.
nico.stehr@zu.de
124 EDUARD BRUCKNER
In general the variations of the Alps' glacial cover correspond with the
weather pattern of their surroundings: glaciers decrease when precipitation
is too low and increase when it is too heavy; in both instances the cause
precedes the effect a little.
Forel and Lang established the same for the temperature.
Undoubtedly the region of the Alps and its immediate surrounding area
experiences long-term variations of precipitation and temperature that are
reflected by changes in the glacial formations. These are typical signs of
secular variations of the weather as Lang called the long-term changes
between wet and dry periods, as well as cool and warm ones.
This view received an unexpected confirmation by a small, but, as far as
the methodology of this entire research area is concerned, significant study
carried out by A. Swarowsky, Wien, about the water levels of the
Neusiedler See.315 He demonstrated that the changes in the water levels of
this lake with no outlet and in the glacial areas exhibit a partly striking
parallel pattern. He thereby delivered proof that even lakes without run-offs
are excellent indicators of secular weather changes-a fact that however
had frequently been assumed a priori.
The variations of the weather over long periods of time as demonstrated
for the Alps and the neighbouring region of the Neusiedler See are in all
respects surprising. They seem to be more pronounced than the variations
of rainfall and temperature following the II-year period of sunspot activity;
for they are followed by glacial changes. While an II-year period of glacial
variations has been claimed by Fritz, to the unbiased observer it does not
exist.
We are still in the dark about these variations; we do not know if they
are limited to the regions of the Alps or extend beyond them to the rest of
Europe perhaps, or to even more substantial parts of the globe. We can
make the assumption that something like this occurs because we know that
glaciers do not remain unchanged but often vary in size. But whether such
variations all occur at the same time, or with delays in some regions and
ahead of time in others-we do not know. And finally, we have no
knowledge whatsoever of their cause.
These were the questions which motivated the research recounted on the
following pages. I could have started by collecting all the data about glacial
variations in European and Non-European mountain ranges. But I was
deterred by the consideration that glacial changes are but an inadequate and
sluggish measuring device for secular variations of weather patterns. It is
first of all necessary to compile the statistical data from observations at
many different glaciers of the same mountain range in order to determine
3\5 Swarowsky in his report about the XIIth year of the Association of Geographers of the
University of Vienna, Wien, 1886, p. 18.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 125
the average periods of increase and decrease because neighbouring glaciers
may often react quite differently due to the peculiar feature of their bed and
their location. Therefore, we have only recently through Forel's work
obtained a more detailed picture of the variations of the Alps' glaciers
during the current century. To achieve the same at present for non-
European mountain ranges, had a priori to be considered hopeless. On the
other hand solving this problem seemed easier and more reliable by
applying Swarowsky's method, e.g., by comparing changes in the water
levels of lakes without run-offs with changes in the rainfall and the
temperature of the regions of the Alps. We will begin with the largest of the
lakes without a constant outlet-the Caspian Sea.
I See Chapter 4b.
nico.stehr@zu.de
4.2 PERIODICITY OF CLIMATIC VARIATIONS
DERIVED FROM OBSERVATIONS OF ICE
CONDITIONS ON RIVERS, THE DATE OF
GRAPE HARVEST AND THE FREQUENCY
OF SEVERE WINTERS*
I. Secular variations of ice conditions on rivers. The significance of the ice
cover during Russian winters. Factors affecting the timing of freezing and
breaking. References, comments, and tables concerning the length of ice-free
seasons and the dates of the break-up. Forming of groups. Applying the results
about temperature fluctuations to Russia and Siberia and to the time dating
back to 1700, in part even back to 1560. Increase in the amplitude of the
variations of ice conditions when progressing in a westerly direction, solely
explained by the value of the periodic temperature variation at the time of
freezing and breaking-up. -II. Secular variations of the start of the grape
harvest. Angot's discourse. Additional data in the form of hand-written notes
by Angot, Forel, and Wehrli. Non-meteorological factors which might
influence the grape harvest. Applying the method of finite differences. Data
and tables for 29 stations in France, SW Germany, and Switzerland
1391-1888. Group average and the overall average for all data series.
Interpretation of the tables. Angot was unaware of variations caused by
climatic oscillations. These oscillations are paralleled by variations in
temperature and rainfall. -III. Table of variations in the frequency of
severe winters 800-1775. Comparison with the variations in ice conditions
and the start of the grape harvest. -IV. The average periodicity of climatic
variations. Table of climatic variations from 1000 to 1880. Average
periodicity of 34.8 (±0.7) years. Climatic variations in central Europe as a
local aspect of global climate variations since 1000.
4.2.1 Secular Variations of the River Ice
Records of the opening and closing of bodies of water date back partly to
the year 1700. It was not only the majestic sight of the ice built-up in a
mighty river with its ice floes slowly gliding downstream at one time, and
crashing, shoving, and swirling at another, which inevitably led to the
[Chapter 8: Die Periodizitiit der Klimaschwankungen, abgeleitet auf Grund der
Beobachtungen fiber die Eisverhiiltnisse der Fliisse, fiber das Datum der Weinernte und
die Hiiujigkeit strenger Winter.]
127
nico.stehr@zu.de
128 EDUARD BRUCKNER
desire to keep track of this phenomenon. It was above all invaluable for
trade and commerce to know exactly when a river would freeze-over in the
winter and when it would open up again. How important the ice cover of
the rivers is varies quite substantially from country to country. In the
oceanic climate of Western Europe rivers freeze rarely, and if at all, then in
abnormally severe winters and for a short period of time only. The further
we move toward the Continental East, the longer the ice cover will last and
the more significant it will be. In the European part of Russia and in Siberia
ice provides a safe passage across the river in winter and at the same time
serves as a smooth road for sleighs commuting up and down the river
between settlements. Of course, it restricts shipping and the transportation
of bulk goods to the warmer season. Thus the river's role as a means of
transportation is twofold. The river is released from this role in the fall,
when its new and fragile ice cover interrupts normal boat traffic and is not
yet able to support sleighs and horses, or in the spring when the ice cracks
and the floes start moving. As much as the ice used to facilitate traffic, it
now prevents it. The river communities that during winter and summer
were able to enjoy close neighbourly contacts, have now, at the beginning
of the freeze-up and even more so at the beginning of the melting, become
inaccessible. At such times traffic comes to a halt for large regions of
Russia.
Before we start discussing the observations, we should address the
question of the causes for the onset of the freeze-up and the beginning of
the break-up of the ice. Should both occurrences indeed be regarded as the
direct result of a change in air temperature? This would only hold true to a
certain degree. A river begins to freeze when the temperature of its top
water level drops below zero degrees. The timing of this occurrence
depends not only on the air temperature but also on the river water itself.
The entire water mass must have cooled to 4°C, i.e., to its maximum
density temperature, before the surface temperature can drop that low. The
time this process takes, depends largely on the initial water temperature and
also substantially on the water volume. The same river will freeze sooner,
when its water levels are low than it would when they are high at otherwise
completely comparable conditions.
Water volume has a similar effect on the timing of the break-up of the
winter ice cover. Very seldom is melting the only cause. A rising water
level is in almost all instances a contributing factor. Caused by the melting
of snow and ice upstream, it pushes up against the ice cover cracking it and
breaking it up and, as a result, starting the formation of ice floes. The faster
and stronger the river-water rises, the earlier this occurs. The rise depends
almost entirely on the intensity of the snow's melting process, i.e.,
indirectly on the temperature in the upstream part of the river's catchment.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 129
The amount of snow fallen during the winter, which provides the melt-
water, as well as the thickness of the ice cover which is determined by the
winter temperatures and the protective layer of snow, are secondary factors
only.
Although the point of freezing and melting of the ice cover as well as its
duration is, in this sense, not a true function of the temperature conditions,
the latter playa major role. Therefore, the condition of the river ice must
primarily reflect the conditions of the air temperature. This does not imply
that changes in the timing and duration of the ice cover from one year to the
next follow precisely the same pattern as the annual temperature averages.
The average temperature, as indicated by the condition of the river ice, is
arrived at through a completely different method from the one used to
determine our annual average where each month carries the same weight.
Winter temperatures and even more so summer temperatures are of little
significance in the determination of this average value as opposed to spring
and fall temperatures-as pointed out by Wild. 316 Secular variations in the
duration of the ice cover as well as in the start of the drift of the ice floes
nonetheless follow the same pattern as the secular variations of the annual
temperature averages. This is, with the exception of the time span after
1840, demonstrated in Wild's study as well as in our table below which
shows that on the whole the duration of the winter ice and the average
temperature vary in a parallel pattern. Consequently, data about the
freezing and melting of the rivers can be used to substantiate secular
temperature variations where actual temperature readings are unavailable.
As an indicator of the temperature, river ice has the advantage over the
thermometer in that it is almost completely free of human interference.
Factors such as faulty instruments and errors in their positioning which can
easily distort the homogeneity of an observation series, do not apply in this
case. Not until the most recent decades has it become possible to arbitrarily
alter the dates of closure and re-opening of rivers with steamship traffic in
the attempt to keep these rivers open for shipping by using machinery, as
shown in the case of the Elbe River. Under these circumstances the
homogeneity of the data appears to be ensured apart, of course, from
writing and printing errors which may have crept into the original
pUblications.
The following sources were used for the tables below:
1. Rykatschew: Uber den Auf und Zugang der Gewiisser des russischen
Reiches [About the opening and closing of the rivers in the Russian
Empire] II. Supplemental volume to Repertorium fUr Meteorologie. St.
316 Wild: Temperaturverhiiltnisse des Russischen Reiches [Temperature conditions in the
Russian Empire], Supplementary Volume to the Repertorium fUr Meteorologie, St.
Petersburg, 1881, p. 285 f.
nico.stehr@zu.de
130 EDUARD BRUCKNER
Petersburg 1887. Numerous printing errors were corrected prior to
incorporating the numerals into the tables, according to the list of
misprints included in the publication and above all according to
Woeikofs detailed review in the Iswestija of the Imperial Russian
Geographical Society, Volume XXIII.
2. For the Donau, Annalen der Hydrographie und Maritimen
Meteorologie, Volume 1880, p. 477.
3. Heydenreich: Klimatische Verhiiltnisse von Lithauen im Regierungsbe-
zirk Gumbinnen nach den 50jiihrigen Beobachtungen zu Tilsit, [The
climatic conditions of Lithuania in the District of Gumbinnen accor-
ding to the observations in Tilsit over a period of 50 years], Tilsit
1870.
4. Draper: Uber die Eisbedeckung des Hudson, [The Ice Cover of the
Hudson], Zeitschrift fUr Meteorologie, 1874, p. 240.
Included in the tables are data from 32 stations about the average annual
length of time during which the rivers are ice-free, and from 12 stations
about the average opening dates of ice-locked rivers, i.e., for each lustrum.
Incomplete lustra are marked by a period symbol.
I paid special attention to the degree of accuracy to which the lustra
averages reflect secular variations of the ice conditions. For this particular
purpose I examined the ice conditions of two rivers, the Newa at St.
Petersburg and the Dwina at Archangelsk, and their variations from one
year to the next by determining for each year the average of its
neighbouring ten years. The final result turned out in favour of the method
of lustra averages; the latter give indeed precise indications of the
variations. It would take up too much space to reproduce the entire table as
originally planned; since Rykatchev published the lO-year-averages for the
river Dwina in his extensive study, I may be permitted to revert to a small
excerpt pertaining to the river Newa for the years 1801-50. The decades are
mentioned in the heading of the table. The left margin shows the individual
years. The 10-year-averages for each individual year were arrived at by
determining the average of this particular year as well as of the previous
five and the successive four years. The lustra-averages are given for their
middle year.
Information about missing years is supplied in the account below.
Stations are categorized according to their geographical location from East
to West, as in the tables.
Lena 1855; Sselenga 1869 f.; Angara 1795, 1803,54,55; Ob (Barnaul)
1836,37,40,41,43,45,46,49, 65; Irtysch 1811,43-48,50; Bogoslowsk
(Teich) 1846, 55, 74 ff.; Kama (Dedjuchin) 1811, 23, 38-45, 55; Kama
(Perm) 1804-24, 27, 29, 35, 36, 44, 61, 80; Belaja 1833, 44, 65; Ai
(Slatoust) 1863,64, 70, 74, 75; Ssyssola 1816, 43, 55-57, 79, 80; Wyt-
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 131
Table 5B.1. Variations of the Ice-free Periods ofthe Newa at St. Petersburg According to 10-
year Averages and Lustra Averages with Deviations from the Mean 1816/80 (219.5 days)
1800 1810 1820 1830 1840
10-y. 10-y. 10-y. 10-y. IO-y.
Mean Lustra Mean Lustra Mean Lustra Mean Lustra Mean Lustra
-4.2 -9.1 11.0 5.3 -7.7
2 -5.3 -8.9 14.6 4.3 -7.7
3 -5.4 -4.9 -9.5 -2.3 14.4 21.1 3.9 1.3 -5.3 -2.9
4 -8.2 -9.5 16.8 -0.1 -3.3
5 -10.8 -7.1 18.4 -1.7 -3.5
6 -14.6 -3.7 18.6 -2.6 -2.1
7 -18.8 4.6 11.8 -4.0 -4.9
8 -16.6 -16.7 2.9 -11.8 11.4 15.7 -5.7 -2.7 -4.0 -2.1
9 -12.9 4.1 8.6 -7.3 -3.2
10 -9.5 4.7 8.5 -6.2 -4.6
schegda 1837, 38, 41, 42, 53, 56, 57, 61, 80; Ssuchona 1793,94,1803,13,
16-18, 20, 22-24, 26, 28, 29, 32-35, 55-58, 61, 80; Wologda 1829 to
1835,37,43,45-54,56-58 (replaced by 55), 61, 79 f., Dwina 1880; Onega
1800-11, 79 f.; Newa 1709, 12; Diina 1813,25,29,33; Wjatka Slobodskoj)
1879 f.: Wjatka (Wjatka) 1816, 55-58, 62, 64, 65, 79 f.: Kama (Jelabuga)
1826,28,33,34,38,39,41,55,80; Wolga (Ssimbirsk) 1831,79 f.; Wolga
(Ssaratow) 1761,62, 1806,50,57,58,80; Wolga (Astrachan) 1814-27,39,
42,44,80; Dnjepr (Jekaterinoslaw) 1847,61,80; Dnjepr (Kijew) 1845-55,
79 f; Weichsel 1833, 34, 80; Meme1 1870 ff.; Donau 1836, 80; Hudson
1816, 17, 60. --Jenissei 1796, 1803, 04, 9-13, 15, 19, 84, 85;
Tschussowaja 1855-61,65,66; Kama 1786,1804-17,24,35,36,58,61,
80; Ssysso1a 1771, 72, 82-84, 87-89, 91, 93-95, 97-99, 1843, 55-57, 80;
Waga (We1sk) 1806, 32, 33, 55, 79 f.: Waga (Werchowashskij Possad)
1801 (replaced by 1800),04,07,32-35,39 f.; Lickscha 1827,30,31,47,
48,57,64; Kyro 1741,43,1851; Abo 1780, 1832,40; Kumo 1828,71,74;
Diina 1559-61, 69, 70, 73-75, 99, 1600,03-08, 10, 11, 13, 14,20,24,25,
between 1640 and 1700 only the following years are available: 1643, 49,
50,53,59,62,67,87,89, and 92.
The lustra-averages of the ice-free periods were expressed as deviations
from the mean of 1816-80 whereas the ones of the rivers' opening dates
were expressed by the difference between the observed lustrum-average
and the mean value of 1816-80. As a result, a negative symbol (-) indicates
that temperatures are too low, a positive symbol (+) that they are too high.
In cases where the 65-year average of 1816-80 could not be obtained
directly it was determined by reduction according to neighbouring stations
of the same group (see below). The use of this method is endorsed by
nico.stehr@zu.de
132 EDUARD BRUCKNER
Woeikof, whereas Rykatschew opposes it. 317 All figures for the break-up of
the ice in Finland's rivers refer to the 65-year average of 1781-1845. The
date is determined according to the new style. Rykatschew mentions that
when he switched from the old to the new style, he neglected to take into
consideration that in the 18th century the difference was not 12 days as is
the case in the 19th century, but only 11. I have ignored this minor
deviation; strictly speaking, however, the times of the opening of rivers in
the past century ought to be reduced by one day, and the smoothed figures
in the table below to be increased by one day. For the 16th and 17th
century-only data from Riga can be traced back that far-this correction
would amount to +3.0 and +2 days respectively. This, of course, has no
bearing on the length of the ice-free period.
Length of Ice-free Time at 32 River Stations
Given as Deviations (Days) from the Mean 1816-80
River Lena Sselenga Angara lenissei Ob Irtysch
Station Kirensk Sselenginsk Irkutsk lenisseisk Bamaul Tobolsk
N. Latitude 57.8 51.1 52.3 58.4 53.3 58.2
E. Longitude 108.1 106.9 104.3 92.1 83.8 68.2
Mean 1816--80 163.9 207.4 279.7 196.8 198.7 189.7
1736-40 -10.3*
41-45 -6.9
46--50 2.3
51-55 12.7
56-60 3.7 9.5
61-65 -3.1 -8.5*
66--70 4.1 -6.3
71-75 -1.7 12.3
76--80 -2.7
81-85 2.1
86-90 -7.5* -4.3*
91-95 5.5 -3.3
96-1800 -5.7 7.3
1801-05 -2.9
06--10 -0.5
11-15 -24.9* -12.9*
16-20 -2.3 -4.1 -8.5
21-25 -1.9 7.1 5.3
1826--30 -3.5 12.3 5.6 2.5
31-35 -7.9* 7.5 -5.6 0.5
36-40 -0.5 -15.5 -10.0* -1.5*
}-12.9*
41-45 2.3 -5.9 0.8
} 2.0
46--50 -5.9 -11.2* -4.3 -3.6 -9.0
51-55 -2.1 0.4 -18.7* -2.6 -10.1
317 Woeikofin the Iswestija of the Imperial Russian Geographical Society Volume XXIII.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 133
River Lena Sselenga Angara lenissei Ob Irtysch
Station Kirensk Sselenginsk Irkutsk lenisseisk Bamaul Tobolsk
56--60 6.8 0.6 4.3 4.9
61-65 2.4 -1.2 -5.9*
66-70 6.6 12.0 0.3
71-75 0.2 -7.5
76-80 1.4
River Ob Teich Kama Kama Belaja Ai
Station Obdorsk Bogoslowsk Dedjuchin Perm Ufa Slatoust
N. Latitude 66.5 59.8 59.4 58.0 54.7 55.2
E. Longitude 66.6 60.0 56.6 56.3 56.0 59.7
Mean 1816-80 148.5 162.0 187.0 208.1 210.8 174.1
1736-40
41-45
46-50
51-60
56--60
61-65
66-70
71-75
76-80
81-85 -18.0*
86-90 -4.4
91-95 -6.4
96-00 -3.3 2.6
1801-05 7.9 -5.4
06-10 -5.4
11-15 0.2 -18.8*
16-20 -5.6 -3.6
21-25 10.9 -8.6
26-30 -3.6 6.9 3.4
31-35 -6.6 1.1 -4.8*
36-40 -2.1 -6.3* -4.2
41-45 -4.5 -6.8* -4.9 0.2 -12.5*
46-50 -13.6* -3.5 -1.0 -2.7 1.2 -12.5
51-55 1.9 3.5 -1.8 -4.9 -0.8 1.9
56--60 3.1 0.0 -7.3* 2.7
61-65 -16.6* -1.8 -11.8*
66-70 3.2 0.3 13.4
71-75 10.3 -0.1 10.2
76-80 -4.1
River Ssyssola Wy!schegda Sschona Wologda Dwina Onega
Station Ust- Ssolwyt- Welikij- Wologda Arch- Onega
Ssyssolsk schegods Ustjug angelsk
N. Latitude 61.7 61.3 60.8 59.2 64.5 63.9
E. Longitude 50.8 46.9 46.3 39.9 40.5 38.1
Mean 1816-80 186.0 192.3 196.3 201.7 179.0 199.0
1736-40 -11.4*
41-45 -3.2
46-50 -11.0
51-55 8.8
56--60 -16.8*
61-65 10.1 -1.0
66-70 9.3 -0.2
nico.stehr@zu.de
134 EDUARD BRUCKNER
River Ssyssola Wy!schegda Sschona Wologda Dwina Onega
Station Ust- Ssolwyt- Welikij- Wologda Arch- Onega
Ssyssolsk schegods Ustjug angelsk
N. Latitude 61.7 61.3 60.8 59.2 64.5 63.9
E. Longitude 50.8 46.9 46.3 39.9 40.5 38.1
71-75 -0.1 * 16.8
76-80 2.5 -5.2
81-85 16.9 -0.0 -8.6
86-90 1.5 -4.4 -3.8
91-95 4.4 -10.4 -13.8*
96-1800 -2.0 -11.2
1801-05 -3.3 -5.4
06-10 -10.1 * -14.7* -14.8*
11-15 -8.1 -13.8 -11.2
16-20 -18.0* -12.4 -13.6*
} -5.4
21-25 -0.6 1.4 6.8
26-30 -1.6 16.6 5.2 16.8
} -3.0
31-35 3.8 1.8 -0.6
36-40 5.4 -4.3 -1.7 0.3 2.0 4.4
41-45 -5.5* -11.6* -11.5* -13.0* -10.2* -20.4*
46-50 6.6 4.9 2.7 1.6 -3.4
51-55 -4.2 7.7 2.2 -0.2 3.8
56-60 5.0 0.7 5.3 2.6 -0.8
61-65 4.4 1.7 0.9 -1.9 -1.8 -0.2
66-70 4.0 2.1 6.3 -0.5 -4.4* 2.0
71-75 1.2 3.3 -5.1* -7.1* -0.8 -6.6*
76-80 -0.2* -0.5* 7.2 1.3 8.4 11.3
River Newa Diina Wjatka Wjatka Kama Wolga Wolga
Siobod-
Station Petersburg Riga Wjatka Jelabuga Ssimbirsk Ssaratow
skoj
N. Latitude 59.9 56.9 58.7 58.6 55.8 54.3 51.5
E. Long. 30.4 24.1 50.2 49.7 52.1 48.4 46.0
Mean
1816-80 219.5 238.0 195.0 203.2 207.1 230.1 235.3
1761-65 7.3 9.7
66-70 2.8 17.3
71-75 -8.3 -6.1
76-80 -5.9 -9.5*
81-85 -3.5 6.7
86-90 -12.2* -1.9
1791-95 5.5 -0.1
96-1800 0.9
1801-05 -4.9 5.2
06-10 -16.7 -5.4* -5.8
11-15 -2.3 -18.0* -1.8 -0.1
16-20 -1l.8 -5.4* -8.5* 8.5
21-25 21.1 33.2 8.0 -1.3
26-30 15.7 7.8 -7.0 10.2
31-35 1.3 10.3 2.2 -2.2 -10.4* -19.6* -17.7*
36-40 -2.7 15.0 0.6 1.6 2.6 -10.5 -13.5
41-45 -2.9 5.6 4.4 -8.2* -3.9 -4.5 2.3
46-50 2.1 14.2 10.6 7.2 -5.1* -5.5 -0.5
51-55 -9.3* -19.0* -1.4 -3.2 2.1 0.5
56-60 -7.7 -5.6 6.2 3.5 -5.6
} 1.6
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 135
River Newa Diina Wjatka Wjatka Kama Wolga Wolga
Station Petersburg Riga Siobod-
Wjatka Jelabuga Ssimbirsk Ssaratow
skoj
61-65 -1.5 0.0 -3.4* 3.3 -0.1 -2.3
66-70 -0.9 -7.6 1.8 2.0 7.1 6.9 6.7
71-75 -7.5 0.0 -2.0 -4.4 -0.8 15.9 10.5
76-80 4.3 7.4 6.7 Il.l 8.9 20.9 9.9
River Wolga Dnje~r Dnje~r Weichsel Memel Donau Hudson
Station Astrachan Jekatarinoslaw Kijew Warschau Tilsit Galatz Albany
N. Lat. 51.5 48.5 50.5 52.3 55.1 45.4 42.7
E. Long. 46.1 35.1 30.5 21.0 21.9 27.9 -73.8
Mean 264.9 277.0 268.7 304.9 264.7 327.8 273.3
1816-80
1781-85 -3.5
86-90 6.1
91-95 19.7
96-1800
1801-05 -14.5*
06-10 2.5 -3.9
11-15 -17.9* -4.3
16-20 -6.7 6.5 -18.9*
21-25 39.4 10.3 12.5 24.5 18.9
26-30 -10.9 2.6 4.5 -20.7 -14.9 10.7
31-35 -7.5 -11.8 -7.3 -20.4 15.7 -5.9
36-40 -2.9 -11.8 -13.9* -24.1* -36.3* -9.4 -7.1*
41-45 6.8 -26.6 5.1 -9.5 ILl -10.8* 1.5
46-50 -1.3 26.0 -8.9 -2.1 -2.8 6.9
51-55 14.1 1.4 13.7 -4.3 -27.2 -8.1*
56-60 2.9 -13.9* -13.7 -19.3 4.8 10.9
61-65 -10.5* -71.6 -ILl -l.l 3.9 -23.8* -3.1
66-70 7.7 7.8 -4.5 14.5 13.5 -1.4
71-75 2.3 -5.2 8.1 2.7 -7.2
76-80 14.9 13.8 -3.9 12.1 7.1
Date of Break-up ofIce in Rivers at 12 Stations
Based on the Differences Between Lustra Avera~es and the Mean of 1816-80
River lenissei Tschussowaja Kama Ssyssola Waga Waga
Werchowashskij
Station leni- Ust-Utkin- Perm Ust- Welsk Possad
sseisk skajaStanika Ssyssolck
N. Latitude 58.5 57.0 58.0 61.7 61.1 60.7
E. Long. 91.1 59.6 56.3 50.8 42.1 42.0
Mean 5. V 21.IV 25.IV 29. IV 25. IV
1816-80 5.Y
1771-75 10.7
76-80 8.8
81-85 } -1.5*
86-90 -2.5
91-95 -8.6*
} 6.3
96-1800 -4.5 -2.0
1801-05 2.0 4.3 -0.4 7.8
06-10 -2.0 -3.6 -6.0* 3.2
11-15 -5.4* -3.0 -5.4*
16-20 2.5 -1.3 -2.8 -2.4 -2.8
21-25 2.6 3.2 1.4 3.2
26-30 0.6 1.2 1.0 4.2 3.0 1.6
31-35 -1.8 -2.6 -2.5 3.0 11.6 4.0
nico.stehr@zu.de
136 EDUARD BRUCKNER
River Jenissei Tschussowaja Waga
Kama Ssyssola Waga
Werchowashskij
Jeni- Ust-Utkin- Perm Ust- Welsk
Station Possad
sseisk skajaStanika Ssyssolck
36-40 -5.0* -3.8 -6.5* 2.4
41-45 3.8 -4.6* -5.6 -3.8* -5.2*
46-50 -3.8 -2.4 0.0 -1.6 5.6
51-55 -0.2 -1.2 0.0 -1.8 5.8
56-60 -1.2 2.8 0.3
61-65 0.0 9.3 -0.2 -1.0 0.2
66-70 3.2 -1.2 -1.0 -2.4 -0.4
71-75 0.4 0.8 0.6 -0.6 2.8
76-80 -0.6 -0.8 -3.2* -4.0* 2.7
81-85 -5.3*
River Lickscha l Boq~al Kyro l Abo 1 Kumo l DUna2
Station Pielis Borga Storkyro Abo Bjomeborg Riga
N. Latitude 63.3 60.4 63.0 60.5 61.5 57.0
E. Longitude 30.l 25.7 22.3 22.3 21.8 24.1
Mean
3. V 24.1V 26. IV 16. IV 24. IV 7. IV
1816-80
1741-45 -1.7 -8.0 -6.6
46-50 4.4 -4.8 0.0
51-55 9.2 1.6 2.6
56-60 1.6 -5.0 2.4
61-65 1.4 0.8 2.6
1766-70 -0.6 -2.8
71-75 0.3 0.6 -4.4 -3.4
76-80 1.3 0.0 -0.2 5.0
81-85 -1.0* 2.0 -7.2
86-90 -0.9 -1.4* -8.6*
91-95 7.9 5.8 5.4
1796-1800 2.1 2.8 -2.8
1801-05 7.8 4.0 3.8 6.0 -2.8
06-10 -10.7* -11.4 -11.0* -11.2* -13.5*
11-15 -0.7 1.8 -0.6 -0.2 -3.4
16-20 -5.1 -5.5 -4.2 -3.6 -4.2
21-25 5.6 2.5 7.2 6.4 11.0
26-30 2.7 0.2 1.I -2.0 0.8 -3.2
31-35 1.0 4.1 3.9 3.2 6.8 11.2
36-40 -1.6 5.8* 0.0 -4.5 -0.2 -6.6*
41-45 -4.2 -l.l -3.3 1.6 -3.8
46-50 8.7 0.6 2.3 -1.2
51-55 56* -4.2 -13.6* -6.2
I) The Finnish stations based on the mean of 1781-1845.
2) Previous years see next table.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 137
River Lickscha l BorSa l K~rol Abo 1 Kumo l Diina2
Station Pielis Borga Storkyro Abo Bjorneborg Riga
56-60 5.4 2.0
61-65 -4.0 4.8
66-70 -3.2 --0.8
71-75 -0.7 -3.2*
76-80 1.2
Date ofIce Break-up of the Diina at Riga I) Ice-free Time of
the Newa
at st. Petersburg 2)
Mean 7. IV I Mean 219.5
+
raw sm. + raw sm. + raw sm.
1556-60 -1.7* -2.1* - -
1561-65 -3.0* -1.0 1687-92 2.7
1566-70 3.7 1.4 - -
1571-75 1.0 1.3 1706-10 - - 9.1 6.2
1576-80 --0.4 0.4 1701-15 1.6 --0.7 0.5 1.2
1581-85 1.6 --0.4 1716-20 -5.4* --0.8* -5.1* -0.5*
1586-90 -4.4 -3.2 1721-25 5.8 0.2 7.7 5.0
1591-95 -5.6 -5.7* 1726-30 -5.4 1.3 9.5 6.8
1596--00 -7.0 -5.5 1731-35 10.2 4.7 0.3 -1.1
1601--05 1736-40 3.8 0.3 -14.5* -8.\
1606-10 -2.4 -2.6 1741-45 -6.6* -2.4* -3.7 -8.8*
1611-15 1746-50 0.0 -1.0 -13.5 -5.5
1616-20 1.5 -1.2 1751-55 2.6 1.9 8.7 1.4
1621-25 -5.3* -2.3 1756-60 2.4 2.5 --0.7 6.1
1641-50 7.3 See tables on pages 110 and 112 for a continuation.
1651-67 -21.4*
In order to eliminate any locally produced particularities individual
observation series were categorised as follows:
For the length of the ice-free period:
1. Siberia: Lena, Sselenga, Angara, River Ob near Barnaul and near
Obdorsk and Irtysk.
2. Ural-Region: Reservoir near Bogoslowsk, Kama near Dedjuchin and near
Perm, Belaja and Ai.
3. North Russia (the entire catchment area of the White Sea): Ssyssola,
Wytschegda, Ssuchona, Wologda, Swina, and Onega.
4. Baltic Provinces: Newa and Diina.
I) Corrections to be added to the lustra averages in order to arrive at the mean value of
multiple years.
2) Deviations from the mean of multiple years.
+ [smoothed]
nico.stehr@zu.de
138 EDUARD BRUCKNER
5. Southeast Russia: Wjatka near Sslobodoskoj and Wjatka, Kama near
Jelabuga and Wolga near Ssimbirsk, Ssaratow and Astrachan.
6. Southwest Russia: Dnjepr near Jekaterinoslaw and near Kijew, Weichsel,
Memel and Donau.
7. Atlantic States, North America: Hudson.
For the dates ofthe ice break-up:
8. Siberia: Jenissei.
9. North Russia: Tschussowaja, Kama, Ssyssola, Waga near Welsk and near
Werchowashskij Possad.
10. Finland: Lickscha, Borga, Kyro, Abo and Kumo.
11. Baltic Provinces: Diina.
Both raw as well as smoothed group averages are included. On page 113
a small table presents the results of the earliest observations of the Newa at
St. Petersburg (1706-60) and the Diina at Riga (1556-1760). They form a
continuous line of data with the group averages of the categories "ice-free
periods in the Baltic provinces" and "dates of the ice break-up in the Baltic
provinces", because the latter's earliest lustra are based on the observations
at St. Petersburg and Riga alone.
Secular Variations of the Length of Time Rivers Remain Ice-free in Different
regions.
Deviations from the Mean (in Da~s) 1816-80: Raw Mean Values.
Lustrum Siberia Ural Northern Russia Baltic Provinces
1736-40 -10.3* -11.4* -14.5*
41-45 -6.9 -3.2 -3.7
46-50 2.3 -11.0 -13.5
51-55 7.8 8.8 8.7
56--{)0 6.6 -16.8* -0.7
61-{)5 -5.8* 4.6 7.3
66--70 -1.1 4.6 2.8
71-75 5.3 8.3 -8.3
76--80 -2.7 -1.4 -5.9
81-85 2.1 -18.0* 2.8 -3.5
86--90 -5.9* -4.4 -2.2 -12.2*
91-95 1.1 -6.4 6.6 5.5
96--1800 0.8 -0.3 -6.6 0.9
1801-{)5 -2.9 1.2 -4.4 -4.9
06--10 -0.5 -5.4 -13.2* -16.7*
II-IS -IS.9* -9.0 -11.0 -10.2
16--20 -5.0 -4.6 -9.7 -11.8
21-25 3.5 9.8 3.3 27.2
26--30 4.2 2.2 6.8 11.8
31-35 -1.4 -3.4 0.5 5.8
36-40 -7.1* -5.2* 1.0 -S.8
41-45 -3.0 -5.0 -12.0* 1.3
46-50 -6.S -3.7 2.5 8.1
51-55 -5.2 -1.8 1.9 -14.2*
56--{)0 3.9 -2.3 2.6 -6.6
61-{)5 -1.6 -10.1 * 0.5 -0.7
66--70 6.3 5.6 1.6 -4.2
71-75 -3.6* 6.8 -2.5* -3.7
76--80 1.4 -4.1 4.6 5.9
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 139
Lustrum SE Russia SWRussia At\. States N. America Means
1736-40 -12.1 *
41-45 -4.6
46-50 -7.4
51-55 8.4
56--60 -3.6
61--65 9.7 4.4
66-70 17.3 5.9
71-75 -6.1 -0.2
76-80 -9.5* -4.9
81-85 6.7 -3.5 -2.2
86-90 -1.9 6.1 -5.1*
91-95 -0.1 19.7 4.4
96-1800 -1.3
1801-05 5.2 -14.5* -3.4
06-10 -2.9 -3.9 -7.1
11-15 --6.6* -4.3 -10.0*
16-20 -1.8 -0.1 -18.9* -7.4
1821-25 3.4 15.8 12.9 10.8
26-30 -2.6 -7.1 10.7 3.7
31-35 -9.2* --6.0 -5.9 -2.8
36-40 -3.7 -19.1* -7.1 -7.1*
41-45 -2.0 -0.8 1.5 -2.9
46-50 0.9 -4.6 6.9 -0.5
51-55 2.0 9.5 -8.1* -2.2
56--60 1.7 -10.5 10.9 0.0
61--65 -1.9 -8.5 -3.1 -3.6
66-70 5.4 6.0 3.5
71-75 3.6 0.4 0.0
76-80 12.1 7.3 4.5
Smoothed Mean Values
Lustrum Siberia Ural N. Russia Baltic Provinces
1736-40 -8.0* -8.7* -8.1
41-45 -5.5 -7.2 -8.8*
46-50 1.4 -4.1 -5.5
51-55 6.1 -2.6 0.8
56--60 3.8 -5.0 3.6
61--65 -1.5 -0.8 4.2
66-70 -0.7 5.5 1.2
71-75 1.7 5.0 -4.9
76-80 0.5 2.1 -5.9
81-85 -1.1 -13.5* 0.5 --6.3*
86-90 -2.2* -8.8 1.2 -5.6
91-95 -0.7 -4.4 1.1 -0.1
96-1800 0.0 -1.4 -2.8 0.6
1801-05 -1.4 -0.8 -7.1 --6.6
06-10 -5.7 -4.7 -10.4 -12.1
11-15 -10.8* -7.2* -11.2* -12.2*
nico.stehr@zu.de
140 EDUARD BRUCKNER
Lustrum Siberia Ural N. Russia Baltic Provinces
16-20 -6.4 -2.2 -6.8 -1.6
21-25 1.6 4.8 0.9 13.6
26-30 2.6 2.7 4.4 14.2
31-35 -1.4 -2.4 2.2 3.7
36-40 -4.6 -4.7 -2.4 -2.6
41-45 -5.0 -4.8* -5.1* 0.5
46-50 -5.5* -3.6 -1.3 0.8
51-55 -3.3 -2.5 ~.2 -6.7
56-60 0.3 -4.1 1.9 -7.0*
61-65 1.8 -4.3 1.3 -3.0
66-70 1.8 2.0 0.3 -3.0
71-75 0.1 3.8 0.3 -1.4
76-80 -0.3 -0.5 2.2 2.7
Lustrum SERussia SWRussia Atl. States N. America Means
1736-40 -9.6*
41-45 -7.2
46-50 -2.8
51-55 1.4
56-60 1.4
61-65 12.2 2.8
66-70 9.6 4.0
71-75 -1.1 0.2
76-80 -4.6* -3.0
81-85 0.5 -0.3 -3.6*
86-90 0.7 7.1 -2.0
91-95 -0.7 15.2 0.6
96-00 -0.4
1801-05 2.5 -11.0* -3.8
06-10 -1.8 -6.6 -6.9
11-15 -4.5* -3.2 -8.6*
16-20 -1.7 2.8 -8.3* -3.5
21-25 0.6 6.1 4.4 4.5
26-30 -2.8 -1.1 7.1 3.8
31-35 -6.2* -9.6 -2.0 -2.2
36-40 -4.7 -11.2* -4.6* -5.0*
41-45 -1.7 -6.3 0.7 -3.1
46-50 0.5 -0.1 1.8 -1.0
51-55 1.8 1.0 0.4 -1.0
56-60 1.0 -5.0 2.6 -1.4
61-65 0.8 -5.4 1.6 -0.9
66-70 3.1 0.8 0.8
71-75 6.2 3.1 2.0
76-80 9.3 4.7 3.0
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 141
Variations of the Date ofIce Break-uE in Rivers in Different Re~ions
Corrections (in Days) Reduced to the Mean 1816-80
Raw Mean Values
Lustrum Siberia N. Russia Finland!) Baltic Provinces
1741-45 -4.8* -6.6*
46-50 -0.2 0.0
51-55 5.4 2.6
56-60 -1.7 2.4
61-65 1.1 2.6
66-70 -0.6 -2.8
71-75 10.7 -1.2* -3.4
76-80 8.8 0.4 5.0
81-85 -1.5 0.5 -7.2
86-90 -2.0* 1.2 -8.6*
91-95 1.2 6.8 5.4
1796-1800 -4.5 2.2 2.4 -2.8
1801-05 2.0 3.9 5.4 -2.8
06-10 -2.0 -2.1 -11.1* -13.5*
11-15 4.6* 0.1 -3.4
16-20 2.5 -2.3 -4.6 -4.2
21-25 2.6 2.6 5.4 11.0
26-30 0.6 2.2 0.6 3.2
31-35 -1.8 2.7 3.8 11.2
36-40 -5.0* -1.0 -2.4 -6.6*
41-45 3.8 -4.8* -1.8 -3.8
46-50 -3.8 0.4 3.9 -1.2
51-55 -0.2 0.7 -7.8* -6.2
56-60 -1.2 1.6 5.4 2.0
61-65 0.0 2.1 -4.0 4.8
66-70 3.2 -1.2 -3.2 -0.8
71-75 0.4 0.9 -0.7 -3.2
76-80 -0.6 -1.3 1.2
81-85 -5.3*
Corrections (in days) reduced to the mean 1816-80
Smoothed mean values
Lustrum Siberia N. Russia Finland Baltic Provinces
1741-45 -3.3 -2.4*
46-50 0.0 -1.0
51-55 2.2 1.9
56-60 0.8 2.5
61-65 0.0 1.2
66-70 -0.3 -1.6
71-75 10.1 -0.6* -1.2
76-80 6.7 0.0 -0.2
81-85 1.0 0.0 -4.5
86-90 -1.1* 1.2 -4.8*
91-95 0.6 3.7 -0.2
1796-1800 -2.3* 2.4 4.2 -0.8
1801-05 -0.6 2.0 0.5 -5.5
06-10 --0.7 -1.2 -4.2* -8.3*
11-15 -3.4* -3.9 --6.1
16-20 2.5 -1.6 -0.9 -0.2
21-25 2.1 1.3 1.7 3.6
26-30 0.5 2.4 2.6 4.0
31-35 -2.0 1.6 1.4 3.2
I) Relative to the mean of 1781-1845.
nico.stehr@zu.de
142 EDUARD BRUCKNER
Corrections (in days) reduced to the mean 1816-80
Smoothed mean values
Lustrum Siberia N. Russia Finland Baltic Provinces
36-40 -2.0* -1.0 -0.7 -1.4
41-45 -0.3 -2.6* -0.5 -3.8*
46-50 -1.0 -0.8 -0.4 -3.1
51-55 -1.4 0.8 -1.6* -2.9
56--60 -0.6 1.5 -0.2 0.6
61--65 0.5 1.2 -1.4 2.7
66-70 -1.7 0.2 -2.8* 0.0
71-75 0.8 -0.2 -1.5 -1.5
76-80 1.5 -0.6 -0.3
81-85 -3.7*
The conclusions that can be drawn from our tables go into two directions.
In the first place, they confirm the findings about the climate variations of
the past century which were based on observations in Western Europe, for
the European and Asian parts of Russia as well and provide, in a most
welcome way, additional data about the development of temperatures dating
back to the year 1700 and even further. It is not until after 1850 that
variations are vague and erratic. This is in accordance with the fact
emphasized earlier, namely that from that point on temperature variations in
general have become much less pronounced and appear to be indistinct. In
contrast, the temperature minimum of 1836-50 is evident everywhere,
regardless whether we look at the length of the ice-free period or the break-
up of the river ice. The latter, in addition, clearly indicates the warm period
around 1860 and signifies the temperature decrease towards the year 1880.
More important are the results concerning more remote periods of time, for
which, generally, very few meteorological observations are available and for
Russia none. Observations at the Diina show a minimum temperature around
1560, followed by a maximum around 1570 and a second minimum around
1595, which is then followed by a maximum around 1616/20: It is note-
worthy that from changes in the Alps' glaciers and of Lake Trasim as well as
Lake Fucin we were able to identify a wet period around 1600, which
corresponds with the cold period observed in Riga at that same time.
Unfortunately from 1626 to 1710 the observations at Riga are incomplete so
that we cannot draw any definite conclusions regarding the variations of the
ice conditions during this time. It seems, however, that there was a cold
period during the twenties of the 17th century, followed by a warm period in
the forties, and again a cold period in the fifties and sixties. Not until the
18th century is it possible again to give an accurate account of the tempera-
ture variations as shown in the following table:
• [See table on page 113]
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 143
Cold Warm
1716-1720 1721-1735
1736-1750 1751-1770
1771-1790 1791-1805
1806-1820 1821-1830
1831-1850 from 1850 on undetermined
This brief presentation corresponds quite well with the one done earlier
on temperature variations. It shows an additional warm period from 1721 to
1735 preceded by a cold period, a fact which was not known to us up to now
for lack of temperature observations, but which is quite consistent with the
water level variations of the lakes. A small inconsistency appears only
during the period of 1750-90. Although both the ice conditions of the rivers
as well as the temperatures suggest that this period was warm at the
beginning and cold at the end, the two factors differ as to the time of this
temperature change: according to the thermometer observations it occurs as
early as around 1755, according to the ice conditions of the rivers not until
around 1770. The latter is exactly the same year that marks the beginning of
the last wet period of the past century, thereby removing the only major
incongruity between the variations of temperature and rainfall which we
encountered earlier. It confirms our assumption that the temperature curve
might have been distorted.
In order to deal statistically with the amplitude of the variations in the
duration of the ice cover as well as in the date of its break-up, I determined
the mean value of the lustra extremes for the different regions based on the
lustra in the tables given above. Only the observations prior to 1855 were
included, while the indistinct development of the last decades was
disregarded.
The figures for the length of time during which rivers were free of ice are
again deviations from the mean value, those for the dates of the ice break-up
are again differences between mean and actual values, so that, throughout,
+ (Plus) means warm and - (minus) cold.
Maximum and Minimum Lustra Averages and the
AmElitudes of the Secular Variations ofIce Conditions on Rivers.
Duration of SE Bait. SW Hud-
Ice-free Time Siberia Ural Russia N. Russia Provo Russia son
Medium of -12.0 -10.8 -8.4 -11.9 -12.5 -16.8 -13.0
Minimum
Medium of
4.3 5.5 8.6 8.0 12.6 15.0 10.7
Maximum
AmElitude 16.3 16.3 17.0 19.1 25.1 31.8 23.7
Date ofIce Break-u£ Siberia N. Russia Finland Bait. Provo
Medium of Minimum -5.0 -3.8 -6.2 -8.1
Medium of Maximum 2.6 5.7 5.9 7.4
Amplitude 7.6 9.5 12.1 15.5
nico.stehr@zu.de
144 EDUARD BRUCKNER
In Siberia the ice break-up in rivers occurs on the average slightly more
than one week later in the coldest lustrum of a secular variation than in the
warmest lustrum, and the ice-free period falls short by 16 days per year.
These differences further increase towards the West. In the Baltic provinces
the rivers open up 15-16 days later in the coldest lustrum than in the
warmest and each winter they remain ice-covered for a full 25 days longer
than in the warmest lustrum. In South West Russia and Poland this amounts
even to a total of 32 days. This geographical trend towards the West is such
a general and consistent phenomenon that its factuality cannot be doubted. It
would, however, be totally wrong to conclude that towards the West the
secular temperature variations increase in size. Rather, the phenomenon has
a completely different source. The freezing and melting of the rivers takes
place approximately around the. time when the annual temperature curve
passes the freezing point. In Siberia, this occurs during seasons with rapid
periodic temperature changes, i.e., in the fall and the spring. Towards the
West this point in time moves further and further into the winter period,
during which periodic changes of the temperature are very gradual. A rise or
fall of the entire temperature curve by a certain amount will therefore in
Siberia cause not more than a slight shift in the times when the freezing
point is passed, whereas in West Russia the effect is considerable. This very
fact is undoubtedly the base of Rykatschew's findings about a decrease in
the variability of the duration of the ice period as well as the freezing and
melting dates, which he claimed to have observed "in an increasing distance
of continental regions from the ocean and also from South to North."318 The
distance to the ocean and the geographical latitude are of equal importance.
Decisive is the extent of the periodic variation in temperature at the time
when melting and freezing usually takes place.
In this way, our tables allow us to draw reliable conclusions about the
existence of temperature variations yet fail to give any indication of their
intensity in different locations. We count 5 112 variations from the beginning
of the past> century up to the year 1885 and can therefore determine an
average duration of 34 years. This result is lower than the one gained from
the shorter observation series pertaining to temperature and rainfall. Yet if
we ignore the years prior to 1740, we arrive at the previous average duration
of 36 years. If we use the observations of the Diina from 1556 on, which are
admittedly quite incomplete, we have a total of 9 variations spanning over
325 years and obtaining again an average of 36 years. These results are fully
supported by the secular variations in the timing of the grape harvest.
318 Rykatchev, op. cit., p. 23
> [18th century]
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 145
4.2.2. Secular Variations of the Time of the Grape Harvest
There seems to be no phenological event whose records date farther back
than the beginning date of the grape harvest. In many regions the individual
wine-grower is not at liberty to begin with the harvest at his own choosing.
An official decree by the administrative authority, be it the mayor or the
local council, is necessary authorizing the start of the harvest and thereby
lifting the ban which was imposed on vineyards some time before the
ripening of the grapes.
Once the starting of the harvest had become an administrative measure,
its execution naturally was put to protocol each year. Records obtained in
this manner, though being sparse, render phenological data, which could not
be more reliable because the release of the date was decided on the advice of
experienced wine-growers.
In fact many old archives and chronicles have already been searched and
scientists have repeatedly undertaken the effort to collect the data.
Consequently, we have today an important body of critically assessed data at
our disposal. Since wine-growing naturally is restricted to certain areas, the
data refer to these regions only. The available records are limited to some
stations in South West Germany and in Switzerland as well as to a large
number of stations in France.
In 1880, when phenological observations were introduced and organized
in France universally, the Bureau Central Meteorologique de France saw it
as one of its duties to collect all the earlier relevant observations.
Automatically, so-to-speak, the first step turned out to be the compilation of
the data about the start of the grape harvest. To this purpose, the
meteorological commissions of the departements were instructed to extract
from the old chronicles and archives all the information pertaining to this
subject. The reports sent to the Bureau were then carefully researched by A.
Angot and their results were published in 1885.319 From this publication we
not only selected the observation series carried out at 19 French stations, the
length of which is remarkable, but also data from Kiimbach in Baden and
Stuttgart. The French records are available in print up to the year 1879. In
addition A. Angot has been kind enough to pass on to me his manuscript
notes about the observations of the year 1880, for which I thank him. For
Switzerland Ch. Dufour's excellent publication320 provided four long
observation series to which are added two published by Angot. I am also
319 Annales du Bureau Central Meteorologique de France. Annee 1883, I. (Paris 1885),
p. B.29 to B.120.
320 Notes sur Ie probleme de la variation du climat, [Note on the problem of climate
variations], Bull. de la Soc. Vaudoise des sc. Naturelles, X., No. 63 (1870), p. 359 fT.
nico.stehr@zu.de
146 EDUARD BRUCKNER
grateful to Prof. Dr. F. A. Forel in Morges for submitting the series of
observations done in Pully near Lausanne and to R. Wehrli in Zurich for the
one from Altstetten.
Consequently, the following results are based on a total of 29 series. For
France I could have easily included 5-10 additional stations from Angot's
publication with observations running over a period of 50-60 years.
However, the series I used provided such reliable and consistent results, that
I saw no further need to increase the amount of data.
Before we enter into a discussion of the tables' contents, we should try to
define the significance of our data. Obviously the timing of the grape harvest
can not serve as a reliable phenological factor in determining climatic
variations unless it is proven that man's interference has little impact and
that changes in the cultivation of grapes did not affect the homogeneity of
the observation series in any major way over time.
The number of factors which, aside from purely meteorological ones,
determine the start of the grape harvest is not exactly small. Those having to
do with the location and conditions of the ground are of no concern to us
here. These may change from location to location, but stay the same in one
site and can therefore never affect the homogeneity of an observation series.
We are only concerned with the factors that for that one site have changed
over time.
The ripening of grapes depends largely on the age of the vine; the grapes
of young vines require a longer time to ripen than those of older vines. But
this fact should be of no importance to us because normally the number of
young vines is very small and grape picking generally will take place
according to the ripening process of grapes from older vines, and not vice
versa. Of course, if in some places diseases like Phylloxera require the
destruction of old vines, which then are replaced by young vines, the
homogeneity of a sequence can definitely be disturbed.
Furthermore, the time it takes for grapes to ripen differs from variety to
variety. In his "Textbook on Agriculture" DeGasparin, using Odart's system,
categorizes all wine varieties into seven groups321 according to the amount of
temperature required for the ripening process. This amount is determined by
using the formerly accepted method of summarizing the temperature
readings from an actinometer" the results for the seven categories are as
follows:
321 Angot, op. cit., p. 34
" [Schwarzirugeithennometer]
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 147
I. 2264° C Grapes
.§ II . 3400
} Medium 3480° C
'O(,)J III. 3564
S N. 4133
'"
'"
..!S V. 4238 } Medium 4250° C
u VI. 4392
VII. 5000
As the categories I and VII have no bearing on the wine regions of
interest to us, and the remaining are grouped around the values of 3480° and
4250° without any major deviations, Angot suggests to distinguish between
two groups only. Actually this is quite sufficient for phenological climate
studies.
The difference between the temperature totals which are necessary to
ripen the grapes of the two groups is quite substantial, and consequently
harvesting takes place at completely different times depending on which
variety is grown in one particular location. Therefore, the introduction of a
new variety of vines could possibly shift the timing of the grape harvest in a
certain region considerably. In the North there is little danger of this ever
happening. Along the latitude of Paris only those grapes will reach maturity
which do not require high temperatures and the introduction of a more
demanding variety is simply impossible. Not so in the South! Here,
naturally, varieties requiring different degrees of temperature are able to
grow side by side. Preference, however, is given to the varieties for a
warmer climate in order to use soil and climate to best advantage. A certain
balance has developed: each region cultivates those varieties best suited to
its particular climate.
Angof22 gives evidence of how conservative the wine-growers of a
certain region are because of these very reasons. Columellus323 describes the
different grape varieties of Gallium. From his description one is clearly able
to distinguish the Pinot grape among other varieties, which even today pro-
duces the highest yield in Burgundy. During the time span for which exact
written records are available, i.e., since the year 1330 for the region of
Beaune and since 1430 for the region of Dijon, viniculture in Burgundy quite
obviously has not changed much: the same variety is still grown, the same
sites still produce the best wine today as they did in the past. Since the time
of Gregor de Tours, i.e., since the VIth century, the same vineyards have
always produced the heavy Burgundy wines. And this is a natural course of
events; because if a vine is transplanted to a location differing, if only
slightly, from the original climatic conditions, the grape loses its distinct
characteristics and the wine its bouquet. All in all it is not likely that a region
322 Angot, op. cit., p.34
323 Angot, op. cit., p. 83.
nico.stehr@zu.de
148 EDUARD BRUCKNER
ever changed its variety on a large scale, a step which would disrupt the
homogeneity of the records for us today.
Cultivation methods are another major factor in the ripening process of
the grapes and consequently in the timing of the harvest. 324 One aspect is for
instance the distance between the planted vines. The closer together they
grow, the more shade there is in the vineyard, the soil remains moist and the
ripening process is delayed. Whether the vineyard is fertilized or not is
another relevant aspect.325 Sulphur, which is often used to fight certain
diseases of the vine, accelerates the ripening process considerably. As an
example, Mares describes 326 the case of the region of Launac, Departement
Herault, where in the years of 1838-54 the vineyards on the average used to
start harvesting on September 19, whereas from 1855 to 1872 when sulphur
was used regularly, they started as early as September 5. Smoke has the
opposite effect, the ripening process is delayed.
In addition to these interferences in connection with various methods of
cultivation, another aspect might induce the wine-grower to start harvesting
prematurely: his fear of field robbers. In this case, according to Angot,327
wholesalers have a vested interest in the picking of the grapes before they
are fully matured because wine made from these grapes will keep longer and
be less susceptible.
Certain weather conditions may be another cause for starting the harvest
before the grapes have reached full maturity. In the eastern and northern
wine regions of France, for instance, cold weather often forces a premature
harvest to avoid a total loss of the crop.328 There have even been years in
which no wine was produced, i.e., in Verdun in 1816, 1821, and 1830.
Last but not least, the timing of the harvest is affected by taste and
custom, which have always and everywhere been subject to changes and
must naturally reflect in our records. But since they occur gradually, they
presumably have no bearing on the rather short-termed variations we are
interested in.
Obviously, the potential sources for error are numerous enough and we
will have to avoid them in order to base our results on solid ground. We need
to take a careful look at our data sequences to ensure their homogeneity and
to discover any infractions. It is quite useful in this regard that of all the
factors influencing the beginning of the grape harvest the meteorological
factors are the only ones which affect large areas in a general way. All that is
needed for the discovery of distorting factors is a comparison between the
324 Angot, op. cit., p.34
325 Dufour, in Bull. Soc. Vaud. des sc. nat., X., p. 396.
326 Ladray; Cours de Viticulture et d'Oenologie, Tome I, p. 594. (Citat d' Angot.)
327 Angot, op. cit., p. 35.
328 Angot, op. cit., p. 113.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 149
data series of neighbouring regions with regard to any changes in the timing
of the grape harvest from one year to the next. Once again it is the method of
determining the differences which allows us to distinguish between suitable
and unsuitable data. This way it was possible to identify the data series from
Perpignan, Pierrefeu near Toulon, Montmorency and Chatillon-sur-Saone as
unsuitable.
Having sorted through the data, I formed lustra averages, which are
shown in the following table. A lustrum average followed by a period
symbol means that this lustrum is incomplete.
Determining the lustra averages was quite simple. In all his tables Angot
defines the beginning of the grape harvest in relation to the days of the
month of September, e.g., the number 30 means September 30, 31 means
October 1,45 means October 15, -2 means August 29, etc. All I needed to
do was to determine the lustra averages of the dates published by Angot in
order to arrive at the average starting date expressed through a day of the
month of September.
These lustra averages are rendered in full days only. The stations are
arranged according to regions and within these according to their
geographical latitude.
With regard to the different stations I would like to offer the following
preceding comments:
For Verdun the year 1803 is missing (1816, 21, and 30 no grapes were
harvested, because they did not reach maturity; the lustra concerned are in
brackets and not included in determining the minimum). For Argenteuil
1751, 1803; Foug 1796, 1804-09,38,40,41,53; Loches 1804,09; Denain-
villiers und Boesses 1740,60, 74, 75, 92, 96 (between 1741 and 1780 figures
refer to the former location, from 1781 on to the latter; Boesses is only 15
kilometres east of Denainvilliers, so that the data from both places can be
integrated); Auxerre 1799; Vendome 1800; Vesoul 1815; Dijon 1391, 95,
96, 1404,06,38,46,47,60,62,92, 1513,21,24,26,29,31,41-44,62, 71,
1650,1720,94-95,1814 (longest series, observations start as early as 1366,
are however up to the year 1390 so fragmentary, that they seem useless for
our purpose); Salins 1563-65, 67, 73, 75, 81, 86, 95, 96, 1605, 10, 15, 18,
19,29,34,35,53, 56, 57, 73, 74, 1709, 10,41 (during the years 1794, 1809
and 1843 no grapes were harvested; the lustra averages concerned are
formed from 4-year lustra and are in brackets); Lons-le-Saulnier 1669. 74,
81,84,88,89,91,94-99,1702,03,06, 14, 15, 17-20,25-27,30,31,37-40,
42, 77, 79, 91, 95, 1803; Pichon-Longueville (in the M6doc region) 1751,
53,60, 1793-1800; Medoc (the series represents the entire M6doc region as
published by the chamber of commerce of Bordeaux in 1882, compo Angot,
op. cit, p. 41); Tain 1879; Castres 1801,02, 10.
Southwest Germany and Switzerland
nico.stehr@zu.de
150 EDUARD BRUCKNER
Missing are: For Stuttgart 1768, 1814, 15, 16; Kiimbach 1613, 17, 19,25,
33,38-40,42, 51, 53, 58-50, 63, 73, 74, 78, 89; Altstetten 1611, 12, 16,17,
32,39,41,43,46-48,51,53, 54, 58, 62, 76, 77, 79, 83, 84, 88, 89, 92, 95,
96, 98, 1709, 15, 33 no harvest, 98-99, 1815, 18, 89, 90. For the villages
around Lake Geneve ranked in a north-east to south-westerly direction:
Veytaux 1741,43,46,48,57,58,63,65,67-69, 72,77, 79, 1800,02; Vevey
1800; Lausanne 1500,01,08, 16,24,25,27,32,34-48,56,67,69,81, 82,
84, 85, 87, 89, 92, 93, 1596-1603, 05-16, 18, 21-28, 30-32, 34-55, 57,
63-66,64-74, 76, 79, 80, 82-86, 1720,21; Pully 1889,90: Lavaux 1581, 84,
91, 1617-23, 25, 28, 49, 51, 1753, 64-66, 1837, 50, 53, 54, 57, 59, 60:
Aubonne 1550, 58, 63, 64, 78, 86-90, 92, 99, 1602, 04, 07, 15, 19, 37, 46,
50,59,64,78,84-86,88,92,94,95,97,99,1704,05,07,08, 10, 14, 18,20,
22, 1857; Rolle 1800-{)2, 05.19, 27, 29, 30, 38, 54.
To provide a clearer picture of the variations of the harvest times and to
enable the combining of observations from several stations into group
averages, individual lustrum averages were not reproduced as such but rather
in relation to the 65-year mean of 1816-1880. Again, I opted for the
correction method; therefore, the table's figures indicate by how much the
observed lustra averages have to be corrected in order to obtain the 65-year
mean. As a result, a minus sign indicates that the temperature was too low, a
plus sign that it was too high, or a plus sign tells us that the harvest occurred
too early, a minus sign too late.
Secular Variations in the Timing of the Grape Harvest,
Given as Lustra Averages in Relation to the Mean of 1816-80.
(- = Harvest Too Late, + = Too Early)
Dijon, Cote-D'Or Raw Smoothed
Mean 1816-80 29 29
1391-95 o 2.7
96-00 8 7.0
1401-05 12 8.0
06-10 0* 3.2*
11-15 1 3.5
16-20 12 10.0
21-25 15 12.0
26-30 6 10.2
31-35 14 7.0
36-40 -6. 2.0
41-45 6 0.0
46-50 -6* -2.5*
51-55 -4 -2.5
56-60 4. 1.8
61-65 3. 2.5
66-70 o 3.0
71-75 9 4.5
76-80 0* 2.5
81-85 1 0.8*
86-90 1 1.0
91-95 1 1.5
Secular Variations in the Timing of the Grape Harvest, Given as Lustra Averages in
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 151
Relation to the Mean of 1816-80. {-= Harvest Too Late, + = Too Earlx}
Individual Stations Dijon, Salins Lausanne Lavaux Anbonne
Raw Data Cote-d'Or Jura Waadt Waadt Waadt
Mean 1816-80 29 43 41 44 44
1496--00 -1 7.
1501-05 2 10.
06-10 3 7.
11-15 -{i* 1*
16-20 0 5.
21-25 22 7.
26-30 1. -4 -2.
31-35 -1. 6
36-40 4 15
41-45 -4
46-50 -2
51-55 1 10 -2
56-{)0 8 11. 10.
61-65 -5* O. 8 o.
66-70 1 -{i. -1.* -5*
71-75 O. 13. 6 -2
76-80 -2 0 -{i.
81-85 2 1. 17. 3
86-90 2 2. 9 16
91-95 -2 -5. -3. 2.* -6.
1596-1600 -3* 6* 11 -7.*
1601-05 5 5.
06-10 1 -4.
11-15 3 -4.
16-20 2 7.
21-25 2
26-30 -4* -8.*
31-35 0 -3.
36-40 11
41-45 4 2
46-50 3 -3
51-55 7 2.
56-{)0 1 11.
61~5 7 6
66-70 11 0
71-75 0* -5.* -16*
76-80 11 8 4
81-85 16 9 -3.
Individual Stations Dijon Beaune Volnay Salins Lons-Ie
Raw Data Cote-d'Or Cote-d'Or Cote-d'Or Jura Saulnier Jura
Mean 1816-80 29 30 29 43 33
1601-05 5 5.
06-10 1 -4.
11-15 3 -4.
16-20 2 7.
21-25 2
26-30 -4* -8.*
31-35 0 -3.
36-40 11
41-45 4 2
46-50 3 -3
nico.stehr@zu.de
152 EDUARD BRUCKNER
Individual Stations Dijon Beaune Volnay Salins Lons-Ie
Raw 'Data Cote-d'Or Cote-d'Or Cote-d'Or Jura Saulnier Jura
Mean 1816--80 29 30 29 43 33
51-55 7 2.
56-60 I 11.
61-65 7 6
66-70 11 0
71-75 0'" -5 .... -16'"
76--80 11 8 4
81-85 16 9 -3.
86-90 7 4 -5.
91-95 2 2 -3
1696-1700 -5'" 10 -8*
1701--05 5 9 4 1 -5
06-10 6 11 6 7. -1.
11-15 -2 3* -2* -3 -7.*
16-20 9. 13 13 10
21-25 -2 6 7 0
26-30 6 11 12 12
31-35 1 7 8 4 -5.
1736-1740 -3 2 3 4
Individual Stations Kfunbach Altstetten Lausanne Lavaux Aubonne
Raw Data Baden329 St. Gallen Waadt Waadt Waadt
Mean 1816--80 45 51 41 44 44
1601--05 16 8.
06-10 12 -1.
11-15 o. 5 13 4.
16-20 5. 0'" -4'" 2.
21-25 5. 10 0
26-30 0 3'" 20. -2*
31-35 -4.* 7 12 -1
36-40 21 19 4.
41-45 o. 11 7 1
46-50 1 2 --6 ....
51-55 13. 4 2
56-60 12 10 8 6.
61-65 6. 11 2 -2.
66-70 9 15 4 0
71-75 0.... 2'" -8'" -11'"
76--80 10. 4 -3
81-85 8 9. 0 -12.
86-90 7. 20 -4. 2 -10.
91-95 3* 9. -8 -9
1696-1700 7 7.* -13'" -14 -20....
1701--05 4 10 -10 -12 -18.
06-10 7 15. -5 -8 -4
11-15 7 7. -12 -17* -12.
16-20 11 15 -8. -15 -16.
21-25 4 10 -9. -12 -16.
26-30 8 19 -6 -6 -9
31-35 6 (10) -11 -13 -15
36-40 8 11 -10 -9 -12
329 Average according to the Stuttgart station reduced to the mean of 1766-1830.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 153
Individual Argenteuil Fong Les
Stations Seine- Meurthe- Lochses Riceys Convignon
Raw Data Verdun et-Oise et-Moselle Aube Aube Aube
Mean 36 28 41 34 34 32
1816-80
1741-45
46-50
51-55 -3
56-60 -2
61~5 2
66-70 -11*
71-75 ~
76-80 -4
81-85 9
86-90 -1
91-95 -1
96-1800 -2 1.
1801-05 -4. 2 2. -1. -1
06-10 -5* 4. 3 3 2
11-15 -2 -1 -3 0 0 3
16-20 -1. -5* -7* -8* -8* -8*
21-25 (4) 0 0 1 1 1
26-30 (0) 2 4 1 1 0
31-35 0 5 1 -1 -1 4
36-40 -5 -2 5. -2 -2 -2
41-45 0 -1 1. -1 -1 2
46-50 -1 5 2 1 1 3
51-55 -8* -3* -5* -7* -7* -5*
56-60 -3 1 4 1 1 -1
61~5 9 6 5 6 6 5
66-70 7 4 10 10 10 6
71-75 3 -2 3 -1 -1* -2
76-80 -3.* -7* 1* ~* 2 -7*
Denain- Vendome Vesoul Dijon Beaune
Individual Stations villiers Auxerre Loire-et- Haute- Cote- Cote-
Raw Data Loiret Yonne Cher Saone d'Or d'Or
Mean 1816-80 29 32 30 24 29 30
1741-45 -2* -5* 2*
46-50 1 0 2
51-55 2 -5 -3 3
56-60 4 -1 2 4
61~5 5 1 1 9
66-70 -8* -12* -7* -1*
71-75 -5. -4 -1 4
76-80 0 -5 -1 5
81-85 Boesses 2 5 10
86-90 2. ~ -2 2
91-95 1. 0 1 6
1796-1800 -3 -2 -1 -3
1801-05 -2 -3 0 5 -2 -4
06-10 1 -3 3 7 -5 -2
11-15 -1 -10* -1 2 -1 -1
16-20 -7* -8 -5* -7* -11* -7*
21-25 3 -1 0 4 -7 -1
26-30 -1 -2 2 1 ~ -3
31-35 3 1 3 4 -3 1
nico.stehr@zu.de
154 EDUARD BRUCKNER
Denain- Vendome Vesoul Dijon Beaune
Individual Stations villiers Auxerre Loire-et- Haute- Cote- Cote-
Raw Data Loiret Yonne Cher Saone d'Or d'Or
Mean 1816-80 29 32 30 24 29 30
36-40 -I -2 -I -2 4 -5*
41-45 -1 -1 -3 0 1 I
46-50 2 2 3 I 2 2
51-55 -6* -6* -9* -8* -3* 0
56-60 4 3 4 1 3 4
61-65 7 6 9 8 9 9
66-70 4 8 3 7 8 8
71-75 2 4 2 0 1 -6*
76-80 -5* -3* -7* -8* -3* 2
Individual Lons-Ie Pichon-
Stations Volnay o Salins Saulnier Longueville Medoc Tain
Raw Data Cote-d'Or Jura Jura Gironde Girone Drome
Mean 1816-80 29 43 33 23 23 28
1741-45 2* 0* -2
46-50 2* 5 -3
51-55 3 3 -9* -10*
56-60 3 6 -2 -9
61-65 9 7 -3 -3
66-70 -2* -6* -12* -11*
71-75 4 0 -4 -5
76-80 8 -1. -8
81-85 10 11 3 4
86-90 3 -4 -3
91-95 6 (10) 5.
1796-1800 -2 3 -3 -4
1801---{)5 -2 5 -3 2 4
06-10 -2 (8) 0 3 3 -2
11-15 -1 3 -I -16* -2 -6*
16-20 -8* -2* -5* -6 -8* -4
21-25 -1 4 2 3 1 -2
26-30 -3 2 1 6 4 -2
31-35 1 2 5 7 5 0
36-40 -5* -3 2 -2 -1 1
41-45 -1 (2) 2 -5 -2 -7
46-50 2 -1 4 2 -1
51-55 -3 -7* -4* -7* -10* -11*
56-60 4 0 3 -1 -1 -2
61-65 7 6 8 4 5 11
66-70 8 5 6 10 9 11
71-75 2 0 -9 2 3 7
76-80 -6* -8* -9* -7* -2* -6*
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 155
Individual
Stations Castres Stuttgart Kiimbach Altstetten Veytanx Vevey
Raw Data Tam Wiirttemberg Baden St. Gallen Waadt Waadt
Mean
28 45 330 45 331 51 45 45
1816-60
1741--45 4* 7* -3*
46-50 8 15 5
51-55 4* 8 0
56-60 7 11 8
61-65 8 10 14
66-70 -1.* 0* 2*
71-75 3 3 6 1
76-80 2 3 10 1
81-85 5 5 13 8
86-90 2 1 10 2
91-95 4 4 5 10
1796-1800 -2 3 2 3.
1801--05 0 -1 -1 2 3 3
06-10 -7 -1 4 4 1
11-15 -6 --4.* -1 0 -5
16-20 -11 * -3. -3 -3* -9*
21-25 0 -3 1 1 0
26-30 -2 -2 -7* 2 1
31-35 -1 3 1 1
36--40 -6* -5* -5 --4*
41--45 0 2 -2 -2
46-50 2 1 -1 3
51-55 --4 -3 -6* -3
56-60 1 1 --4 1
61-65 7 5 5 7
66-70 8 4 4
71-75 3 5 4
76-80 0* -2 -4.*
81-85 1
86-90 -8*
Individual Stations Lausanne Pully pres Lavaux Aubonne Rolle
Raw Data Waadt Lausanne Waadt Waadt Waadt Waadt
Mean 1816-80 41 44 44 42 41
1741-45 -16* -13 -20*
46-50 -9 -13 -16
51-55 -14 -15.'" -18
56-60 -9 -7 -13
61-65 -6 0 -12
66-70 -14* -18.'" -16'"
71-75 -8 -9 -12
76-80 -2 0 -5
81-85 -3 1 -1
86-90 -5 -2 -9
91-95 7 7 3
1796-1800 -5 -2 -5
1801--05 0 -9
330 mean 1766-1830
331 Mean according to the Stuttgart station mean of 1766-1830.
nico.stehr@zu.de
156 EDUARD BRUCKNER
Individual Stations Lausanne Pullypres Lavaux Aubonne Rolle
Raw Data Waadt Lausanne Waadt Waadt Waadt Waadt
Mean 1816-80 41 44 44 42 41
06-10 -1 -1 -7 -4
11-15 -7 -6'" -10'" -6
16-20 -II.'" -5'" -6 -10 -10.'"
21-25 -4 -4 2 -I 4
26-30 -2 -2 0 0
31-35 -1 -1 -1 2. 2
36-40 -3 -3 -3'" -3 -1.
41-45 -6'" -6'" 2 -4 -1
46-50 4 4 3 0 2
51-55 -5. -5 -2. -5'" -6.'"
56-60 0 0 -2. 3
61-65 9 8 6 9 8
66-70 5 6
71-75 3 4
76-80 0 -6.'"
81-85 0
86-88 -2'"
Overall Avera~es
Raw Smoothed Raw Smoothed
1496-00 3.0 3.2
1501-05 6.0 5.0 1601-05 8.5 4.4
06-10 5.0 304 06-10 2.0 4.0
11-15 -2.5'" 0.6'" 11-15 3.5 2.8
16-20 2.5 4.2 16-20 2.0 2.8
21-25 14.5 7.4 21-25 3.8 2.8
26-30 -1.7 304 26-30 1.5'" 2.2'"
31-35 2.5 3.2 31-35 1.8 4.7
36-40 9.5 4.4 36-40 13.8 8.4
41-45 -4'" 0.1 41-45 4.2 5.4
46-50 -2 -1.2'" 46-50 -0.6'" 2.2'"
51-55 3.0 3.4 51-55 5.6 4.7
56-60 9.7 5.8 56-60 8.0 6.6
61-65 0.8 2.1 61-65 5.0 6.1
66-70 -0.2'" -0.2'" 66-70 6.5 3.2
71-75 4.2 0.7 71-75 -504'" 004'"
76-80 -2.7 1.2 76-80 5.7 3.8
81-85 5.8 4.0 81-85 4.0 4.1
86-90 7.2 4.4 86-90 2.6 2.2
91-95 -2.8'" 0.1'" 91-95 -0.6 -0.8
1596-1600 -1.2 0.9 1696-1700 -4.5'" -2.7'"
SWGermany
Raw Data North France Central France and Switzerland Mean
1701-05 2.8 -5.2 -1.2
06-10 5.8 1.0 304
11-15 -2.2 -5.4 -3.8
16-20 11.2 -2.6 3.6
21-25 204 -4.6 -1.1
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 157
Raw Data North France Central France SWGennany Mean
and Switzerland
26-30 10.3 1.2 5.2
31-35 3.0 -4.6 -0.8
36-40 1.5 -2.4 -0.7
41-45 -2.0* -0.6 -6.8* -3.8*
46-50 1.0 1.2 1.7 0.2
51-55 -2.0* -2.2* -5.8 -3.6
56-60 0.3 0.7 -0.5 0.1
61-65 2.7 3.3 2.3 2.8
66-70 -10.3* -6.5* -7.8 -7.8*
71-75 -5.0 0.3 -2.3 -2.1
76-80 -3.0 0.7 1.3 0.9
81-85 5.5 7.2 4.0 5.5
86-90 -1.7 -0.5 -0.1 -0.6
91-95 0.0 5.6 5.7 4.5
1796-1800 -1.5 -1.7 -0.7 -1.3
1801-05 -0.9 0.6 -0.3 -0.2
06-10 1.0 3.0 -0.6 0.2
11-15 -1.7 -2.9 -4.9 -3.1
16-20 -6.3* -6.9* -6.7* -6.6*
21-25 1.0 0.3 -0.4 0.3
26-30 0.8 -0.2 -1.2 -0.2
31-35 1.7 2.1 0.8 1.5
26-40 -1.3 -1.7 -3.4 -2.1
41-45 -0.6 -0.9 -1.4 -0.9
46-50 2.0 1.4 2.0 1.8
51-55 -6.2* -5.7* -4.4* -5.5*
56-60 1.6 1.2 -0.1 1.0
1861-65 6.6 7.4 7.1 7.0
66-70 6.9 8.0 4.8 7.0
71-75 0.9 1.1 4.0 1.5
76-80 -3.9* -5.5* -3.0 -4.4
81-85 0.5 0.5
86-88 -5.0* -5.0*
Smoothed SW Gennany and Total
data North France Central France Switzerland Average
1701-05 3.8 -3.1 -0.9
06-10 3.0 -2.2 0.4
11-15 3.2 -3.1 -0.2
16-20 5.6 -3.8 0.6
21-25 6.6 -2.6 1.6
26-30 6.5 -1.7 2.1
31-35 4.4 -2.6 0.7
36-40 1.4 -4.0 -1.5
41-45 -1.0* 0.4 -4.4* -2.1*
46-50 0.5 -0.1 -4.0 -2.0
51-55 -0.7 -0.6* -3.4 -1.8
56-60 0.3 0.6 -1.1 -0.2
61-65 -1.2 0.2 -0.9 -0.5
66-70 -5.7 -2.5* -3.9* -3.7*
71-75 -5.8* 1.6 -2.8 -2.8
76-80 -1.4 2.1 1.1 1.3
1781-85 1.6 3.6 2.3 2.8
86-90 0.5 3.0 2.4 2.2
91-95 -0.8 2.2 2.6 1.8
1796-1800 -1.0 0.7 1.0 0.4
1801-05 -0.6 0.6 -0.5 -0.4
06-10 -0.2 0.9 -1.6 -0.7
11-15 -2.2 -2.4 -4.3 -3.2
16-20 -3.3* -4.1* -4.7* -4.0*
nico.stehr@zu.de
158 EDUARD BRUCKNER
Smoothed North France Central France SW Germany and total
data Switzerland Average
21-25 -0.9 -1.6 -2.2 -1.6
26-30 1.1 0.5 -0.5 0.4
31-35 0.7 0.6 -0.8 0.2
26-40 -0.4 -0.6 -1.8* -0.9
41-45 -0.1 -0.5 -1.0 -0.5
46-50 -0.7 -1.0 -0.4 -0.7
51-55 -2.2* -2.2* -1.7* -2.0*
56-60 0.9 1.0 0.6 0.9
61---(i5 5.4 6.0 4.7 5.5
66-70 5.3 6.1 5.2 5.6
71-75 1.2 1.2 2.4 1.4
76-80 1.2 -3.3* -0.4 -1.7
81-85 -1.8 -2.1
86-1888 -3.2* -3.2*
To eliminate any remaining chance occurrences from the lustra averages
of individual stations, I grouped regions together as follows and formed
averages for these groups for the period from 1701-.:.1885 [on pages 132 and
133].
1 . Northern France: Verdun, Argenteuil, Foug, Loches, les Riceys,
Couvignon, Denainvilliers, Auxerre und Vendome. 9 Stations.
2. Southern France Vesoul, Dijon, Beaune, Volnay, Salins. Lons-Ie-
Saulniers, Pichon-Longueville, M6doc, Tain und Castres. 10 Stations.
3 .South West Germany and Switzerland: Stuttgart, Kiirnbach,
Altstetten,Veytaux, Vevey, Lausanne, Pully, Lavaux, Aubonne u. Rolle.
10 Stations.
In addition from the start one single group average was determined for
the observations from all stations [on page 132].
The tables show for all group averages raw figures as well as those which
were smoothed according to the method I have repeatedly used earlier.
Let us discuss our tables:
Each single observation series tells us that the date of the grape harvest
changes not only from year to year but also from lustrum to lustrum. These
are not random changes but show a certain pattern; at times the harvest is
delayed longer and longer over several lustra, only to start earlier and earlier
in the succeeding lustra. There are distinct periods when harvesting started
very late and those when the crops were brought home quite early.
It is remarkable that all these variations do indeed occur simultaneously
at all stations from northern France to the South, from the West to
Wiirttemberg and into Switzerland.
The lustrum of 1876/80 shows that, with one exception, harvesting
started later than in the previous lustra at all stations, the same goes for the
lustrum of 1851155. Between these two "minima" lies a clearly pronounced
"maximum" for all the stations half of which show it in the lustrum of
1861165, the other half in 1866/70. During this time grapes were on the
average picked 1 112 to 2 weeks earlier than in the lustra of 1851155 and
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 159
1876/80. An analogous maximum occurs around the year 1830. Differing
slightly from station to station it appears in five instances as early as in the
lustrum of 1821125, in four instances in 1826/30, ip. 13 in 1831135 and in
three in 1836/40. The years 1836 to 1855 are generally marked by late
harvests; but in the second lustrum of the 1840s a slight return to an earlier
date emerges which is more pronounced at some stations so that the absolute
maximum prior to 1851155 falls into this lustrum. Its cause appears to be a
local one, since it does not show up at all in the smoothed group averages
where the lustra of 1826/30 and 1831/35 represent the earliest time of the
harvest. - The year 1815 is surrounded by minima again, which at one
station appear in the lustrum of 1806/10, at six stations in 1811115, and at 20
in 1816/20. Ifwe return to the previous century, the lustra from 1781 to 1795
show maxima, 1766170 a minimum at all stations, 1756-1765 a maximum,
1741156 again a minimum which seems to have been a general occurrence
and was preceded by a vaguely defined maximum in the years 1716-1730.
The minimum of 1691-1700, clearly identifiable at five stations in
Switzerland and neighbouring France, shifts at four other stations to the
lustrum of 1711115. Moving further back in time, variations tum irregular.
However, some periods of fairly regular patterns still occur: minima around
1671-75, 1626-35, 1591-1600, 1561-70; maxima 1651-60, 1601-05,
1581-90 and 1556-60. Dijon, Salins, Lausanne are the only stations
providing records of the time prior to 1550; for the time before 1495
observations are only available from Dijon. These observation series seem to
indicate that the lustra of 1521125, 1501110, 1471175 and 1421125 carry
maxima, the lustra of 1541/50, 1511115, 1496/1500 or 1476/80, 1446/50 and
1406/10 minima. However, the variations before 1550 should be considered
unreliable due to the low number of stations.
These distinct oscillations have clearly escaped Angot's as well as
Dufour's attention despite the fact that both examined the data series of their
publications with regard to climatic variations. The number of years that
they combined into averages was too big. By using periodicities of 10 or
even 25 years they could not possibly detect variations of under 30 to 40
years. They did however come across variations of the harvest date that
occurred in the course of centuries. For example, from 1550 to 1670 the
harvests at Aubonne took place on the average around October 13, and again
around that same date in the years after 1780. But in the years 1670-1780
the average date was several days after October 20. Lausanne and Veytaux
show analogous and parallel variations over very long periods of time. The
data series from Salins and Dijon again indicate such variations, which
however neither seem to be in accord with each other nor with the variations
in Switzerland. Angoe 32 comes to the conclusion that they may not be
332 Angot, see B83.
nico.stehr@zu.de
160 EDUARD BRUCKNER
attributable to climate variations but rather to changes instigated by hUp1an
arbitrariness, be it a change in taste or in the methods of cultivation. To what
extent this conclusion is justified will have to be discussed later on.
As opposed to long-term variations our variations over shorter periods of
time have the advantage of occurring universally. They reflect the climate
variations that we have already observed many times before.
In the last part of his essay Angot has dealt in detail with the question of
the meteorological factors which mainly determine the date of the grape
harvest. In order to investigate this question for France, he concentrated on
the years 1811, 1822, 1834, 1846, 1865, and 1868 showing very early
harvest dates, and the years 1816, 1821 and 1879 with very late harvests.
Based on the Paris observations he determined the extent to which
temperature and rainfall deviate from the standard during the vine's growing
period. His results are summarised below:
Deviations from the Paris Standard:
Temperature Temperature- Date of the
April to August sum Grape Harvest
Early Harvest + 1.39° C +212° C 17 days too early
Late Harvest -1.52 -223 20 days too late
Negative temperature deviations during the vegetation period obviously
cause a delay of the harvest, whereas positive deviations make early harvests
possible. During the years in question, however, rainfall shows only small
deviations from the standard amount.
With this result we are already in a position to link those changing
harvest dates to climate variations. The variations in the average temperature
during the vines' growing season have without doubt caused those changes
in the harvest dates. Average temperatures for the months April to
September are not at my disposal in order to further support this conclusion.
Yet, these monthly averages are included in the annual averages, the
variations of which we have already examined earlier. In the following table
the variations of the harvest dates are therefore compared with changes of
the annual temperature in Western Central Europe and of the rainfall in
Southern Germany and France (averages of the observation series from
Southern Germany, Northern, Southwest, and Central France).
Congruency between the developments of the three factors-date of the
grape harvest, temperature, and rainfall-is excellent during the present
century; apart from slight shifts in timing, the variations correspond in such a
way that an early harvest coincides with high temperatures and low rainfall,
and vice versa). For the past century, congruency between temperature and
grape harvest is again quite satisfactory while rainfall shows deviations in
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 161
the period of 1766-1790: its minimum in 1766170, as well as its maximum
in 1786/90 are not congruent.
I do not want to attach too much importance to this result. If we, for
instance, replace the less than reliable rainfall observations with those of the
Alps' glaciers, the Neusiedler See and the Caspian Sea, it is evident that
even during the time period in question the changes in the harvest dates must
have been accompanied by corresponding changes in the rainfall. The
parallelism between the variations of the harvest dates and the temperature is
clearly established for the period 1756 to 1875, e.g., for a total of 120 years,
and between harvest dates and rainfall for the period 1731 to 1880, e.g., for
150 years.
This parallelism is so complete that the two curves are interchangeable.
The variations of the grape harvest dates reflect the climate variations as we
have confirmed them, just as well as temperature and rainfall. This finding is
of the greatest importance to us because evidently the curve of the grape
harvest dates can now serve as an indicator of climatic variations for even
those periods reaching far back in time when temperature and rainfall
observations were not available. We are able to trace climatic variations
back to the year 1400 thanks to the many extensive listings of the harvest
dates in the wine-growing regions of France and Switzerland.
For those time periods dating back even further, a somewhat different
kind of data presents itself, which may not be quite as exact but will
nevertheless supply useful results: the records about cold winters
Secular Variations of the Grape Harvest Dates, Compared with the Variations
of Temperature and Rainfall
Lustrum GraQe-harvest Days TemQerature {oq Rainfall {%2
1731-35 -0.8 7
36-40 0.7 14
41-45 -3.8* -3
46-50 -0.2 4
51-55 -3.6 -12*
56-60 0.1 -0.52 5
61-65 2.8 -0.53 0
66-70 -7.8* -0.98* -13*
71-75 -2.1 0.00
76-80 0.9 0.36 -5
81-85 5.5 -0.10 -5
86-90 -0.6 -0.09 7
91-95 4.5 0.48 0
1796-1800 -1.3 0.14 -4*
1801-05 -0.2 0.04 2
06-10 0.2 0.07 7
11-15 -3.1 -0.39* 6
16-20 -6.6* 0.36 -3
21-25 0.3 0.52 -7
nico.stehr@zu.de
162 EDUARD BRUCKNER
Lustrum Grape-harvest Days .1'emperature (DC) Rainfall (%)
26--30 --0.2 --0.23 -I
31-35 1.5 0.35 -8*
36-40 -2.1 --0.30 2
41-45 --0.9 0.09 8
46--50 1.8 0.05 0
51-55 -5.5* --0.31* 2
56--60 1.0 --0.09 -I
61--65 7.0 0.36 -14*
66--70 7.0 0.31 I
71-75 1.5 --0.14 0
76--80 -4.4* 14
Secular Variations of the Frequency of Severe Winters:
The Number of Severe Winters Among- a Series of 20 Winters
800 2 970 I 1140 5 1310 8* 1480 4 1650 9*
05 3* 75 2 45 6 15 7 85 4* 55 8
10 2 80 2 50 7* 20 5 90 4 60 9*
15 2 85 4 55 7* 25 5 95 4 65 8
20 2 90 5* 60 5 30 4 1500 3 70 4
25 2 95 4 65 5 35 5 05 5 75 4
30 2 1000 4 70 5 40 5 10 5* 80 2
35 I 05 3 75 3 45 4 15 5* 85 6
40 I 10 3 80 4* 50 4 20 5 90 7
45 0 15 3 85 3 55 5 25 3 95 6
50 2 20 3 90 I 60 5* 30 3 1700 8*
55 3 25 I 95 2 65 5 35 3 05 4
60 3 30 0 1200 4 70 4 40 4* 10 4
65 4* 35 1 05 6 75 2 45 4* 15 4
70 3 40 3 10 7* 80 3 50 3 20 4
75 3 45 3 15 7* 85 5 55 5 25 5
80 3 50 5* 20 6 90 6* 60 5 30 4
85 3 55 4 25 5 95 5 65 5 35 5*
90 2 60 4 30 5 1400 6 70 6* 40 4
95 1 65 5 35 4 05 4 75 4 45 4
900 70 6* 40 4 10 3 80 6 50 3
05 75 4 45 3 15 4 85 6 55 2
10 80 3 50 4* 20 3 90 5 60 4
15 1 85 2 55 3 25 7 95 8 65 4
20 3 90 3 60 3 30 7 1600 7 70 5
25 2 95 4 65 4 35 9* 05 8 75 8*
30 2 1100 5 70 4 40 8 10 9 80 -
35 4* 05 6 75 6 45 4 15 10* 85 -
40 2 10 7 80 7* 50 4 20 8 90 -
45 2 15 8* 85 7* 55 1 25 6 95 -
50 2 20 8 90 6 60 1 30 5
55 0 25 8 95 6 65 2 35 3
60 0 30 5 1300 6 70 2 40 4
65 1 35 5 05 6 75 3 45 5
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 163
4.2.3. Secular Variations of the Frequency of Cold Winters
Much has been written about the periodicity of severe winters;333 focussing
in most cases more or less on the individual year. I am not convinced that
this method will lead to any sound results. The severity of winters depends
after all on so many factors that 'chance' has to playa major role in its
occurrence, especially when a single year is concerned. It is completely
inadmissible to conclude from one severe winter that a cold weather period
occurred as we have just done from temperature and ice observations. The
situation is different, when instead on the single year attention is turned
towards a possible periodicity in the frequency of severe winters. Based on
Pilgram's listing of severe winters I have done just that and examined their
frequency with regard to any periodic patterns. Pilgram's essay was not
available to me in the original but in a hand-written excerpt instead, for
which I thank Professor Dr. W. Koppen. From this listing, using five-year
lustra I determined the number of severe winters for every 20 years, e.g., for
the time spans 1741-60, 1746-65, 1751-70, 1756-75, etc., and assigned
these frequencies to the middle year of the corresponding time span, e.g.,
1750, 1755, 1760, 1765, etc. This method resulted in the table [on page 138]
which applies to Central Europe because the observations refer to that region
only.
As far back as we are able to compare the variations in the frequencies of
severe winters with those of the temperature or the ice conditions of the
rivers, the three factors correspond quite well; the temperature minima
around 1770,1740,1660 and 1615-20, as well as the maxima around 1755,
1720-30 and 1680 are all reconfinned. In other time periods congruency is
less clear. The maximum cold period at the beginning of the 18th century
shows up in 1700 and then again in the years 1711-20, another one in 1615
and 1621-25. A third one around 1525 is not more than indicated in the
listing of severe winters, whereas it is quite pronounced at the Diina river.
The time periods around 1570 and 1560 coincide again as expected.
In comparing the variations in the grape harvest dates with those in the
frequencies of severe winters, as shown in the table below (page 142),
congruency again turns out to be satisfactory. I must confess that its high
degree was unexpected. The transition from one period to another often
shifts by five or even ten years here and there: e.g., according to the harvest
dates a cold period is indicated for the years 1436-55, whereas according to
the frequencies of cold winters this period falls into the years 1425-55.
However, most periods correspond. Of the 385 years, for which observations
regarding the frequency of cold winters as well as the grape harvest dates are
333 See Chapter 4a.
nico.stehr@zu.de
164 EDUARD BRUCKNER
available, not less than 260 years or 68 percent show the two factors to be
congruent in indicating a cold or warm period and only 125 years show
differences. Only once do the variations of the two factors differ
considerably: according to the harvest dates the years 1591 to 1690 are
marked by four variations, whereas according to the frequencies of severe
winters this number amounts to only two.
In order to decide which of the two findings is the correct one, we will
compare the raw lustrum averages of the two factors (tables pp. 132 and
138) for the time span in question. It turns out that the warm period of
1601-10 which shows up clearly in the dates of the grape harvest, is vaguely
outlined in the listing of cold winters as well: 1595, 8; 1600, 7; 1605, 8.
Therefore, I am inclined to put the emphasis on the harvest dates in this case.
The cold period of 1666-75 is a different matter. It seems to emerge from
the harvest dates, yet stretches over only 10 years in the smoothed data,
while being restricted to the one lustrum of 1671-75 in the raw data:
1656-60, 8.0; 1661-65, 5.0; 1666-70, 6.5; 1671-75, -5.4; 1676-80, 5.7;
1681-85, 4.0. If we ignore the lustrum in question, the cold period has
disappeared entirely. The data about the frequency of cold winters are,
therefore, decisive in this case. As a result the variations' occurrence during
the time period under discussion is depicted in the table below under the
heading 'corrected'.
According to Date According to Frequency
of the Wine Harvest of Severe Winters Corrected
Cold 1591-00 Cold 1591-00
Wann 1601-10 } Cold 1590-1625 Wann 1601-10
Cold 1611-35 Cold 1611-35
Wann 1636-45 Wann 1626-45 Wann 1636-45
Cold 1646-50
Wann 1651-65 } Cold 1645-65 Cold 1646-65
Cold 1666-75
Wann 1676-90 Wann 1665-85 Wann 1666-85 resp. 90
Given the high degree of congruency between the variations in the
frequency of cold winters and the grape harvest dates during the time period
from 1390 to 1775 there can be no doubt that the former factor is the correct
indicator for the climatic variations. The same seems likely for the years
back to the year 1000. Further back in time, the number of listings for cold
winters decreases so rapidly, and in addition the intervals between two
maxima of the same temperature increase by so much that I distrust these
data. I am inclined to regard those variations as coincidental and will
therefore exclude them from further consideration.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 165
4.2.4 Average Periodicity of Climatic Variations
Meteorological observations allowed us to trace climatic variations from not
earlier than approximately 1730 on. However, the number of 4 112 oscilla-
tions which have occurred since then is too low in order to define the
average periodicity. But since we have collected a larger amount of data
consisting, on the one hand, of observations of ice conditions on rivers from
1556 to 1885 and of the grape harvest dates from l391 to 1888, and on the
other hand of listings about the frequency of severe winters from 1000 to
1775, we can approach this task with a better chance of successfully
completing it. The following table has been set up for this purpose. Based on
the smoothed results of the observations about the start of the grape harvest,
and on statistics about cold winters as well as about the ice conditions of the
rivers it shows the extent of warm and cold periods. To avoid any
arbitrariness in distinguishing between the periods, for the time prior to 1750
I used the method of drawing the line wherever the figures coincide with the
mean of the nearest maximum and minimum. For the time after 1750, in
addition I took temperature observations into account whenever cases were
uncertain. I have also included the data about the changes in lakes with no
outlet and glaciers from chapter III.
The table includes the years 1020 to 1890 (or 1888 to be exact). Within
this time span of 870 years we count 25 cold and 25 warm periods, that is 25
complete variations. We, therefore, arrive at an average periodicity of 34.8
years, which is slightly lower than previously assumed (36 years).
To find out how significant this figure may be, I determined the length of
each variation according to both, the harvest dates and the frequency of cold
winters, as shown in the last two columns of the table (p. 142). I arrived at
these figures by combining the numbers for the duration of each cold period
with those of its neighbouring warm period, that is I summed up the years of
the warm periods from 1020-40 and the cold periods from 1040-55 (20 + 15
= 35), furthermore the cold periods from 1040-55 and the warm periods
from 1055-65 (15 + 10 = 25), etc.
The length of these periods varies between 20 and 50 years, but is most
frequently close to the average value of 34.8 which we established above.
The frequency of these periods of varying length is shown in the following
tabulation:
Period: 20 25 30 35 40 45 50 years
Frequency: 6 10 12 13 12 8 4 cases
This verifies the significance of the average figure of 34.8 years. There are
also other methods to verify this.
nico.stehr@zu.de
166 EDUARD BRUCKNER
Secular Variations of the Climate' Represented by the Variations of
Grape Harvest Frequency of Cold Ice Conditions of Rivers
Winters
Wann 1020-40
Cold 40-55
Wann 55--65
Cold 65-80
Wann 80-05
Cold 1105-30
Wann 30-45
Cold 45--65
Wann 65-75
Cold 75-90
Wann 90-00
Cold 1200-30
Wann 30-45
Cold 45-55
Wann 55-70
Cold 70-90
Wann 90-10
Cold 1310-25
Wann 25-50
Cold 50-70
Wann 70-85
Cold 1391-15 85-05
Wann 1416--35 1405-25
Cold 36--55 25-55
wann 56--80 55-75
cold 81-95 75-95
wann 96--10 95-05
cold 1511-15 1505-20
wann 16--40 20-35
cold 41-50 35--45
wann 51--60 45-55
cold 61-80 55-70 1556--65
wann 81-90 70-90 66--85
cold 1591-00 86-00
wann 1601-10 1601-20
cold 11-35 21-25
wann 36--45
cold 45--65 1651--67
wann 65-90 resp .. 85
cold 1691-05 1685-05 1702-20
wann 1706--35 1705-30 21-35
cold 36--59 30-50 36--50
wann 56--65 50--65 51-70
cold 66--75 65-75 71-90
wann 76-05 91-05
cold 1806--20 1806--20
wann 21-35 21-30
cold 36--55 31--60
wann 56--75 }wann
cold 76--90 }1861-80
[This table as well the those on the next three pages was presented by Brueckner in one big
table; unfortunately, because of spatial limitation, the table had to be broken into four
parts.]
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 167
Secular Variations of the Climate"
Represented by the Length of periods as given
variations of ice conditions by
of
Grape Frequency
Lakes Glaciers Harvest of Cold
Dates Winters
Warm
Cold 35
Warm 25
Cold 25
Warm 40
Cold 50
Warm 40
Cold 35
Warm 30
Cold 25
Warm 25
Cold 40
Warm 45
Cold 25
Warm 25
Cold 35
Warm 40
Cold 35
Warm 40
Cold 45
Warm 35
Cold 35
Warm 45 40
Cold 40 50
Warm 45 50
Cold 40 40
Warm 39 30
Cold 20 25
Warm 30 30
Cold 35 25
Warm 20 20
Cold 30 25
Warm 30 35
Cold 1600 I) 1595-10 20
Warm 20
Cold 38 2) 35
Warm 56 3) Increase 35
[continued on next page}
.. [This table was part of a much larger table in the Brueckner's dissertation. It included the
table presented on the previous two pages plus the continuation of this one on the next
page.]
1) Lake Fucin and Trasimen.
2) Caspian Sea, at high tide, compared with levels of 1715/20.
3) Lake Fucin.
nico.stehr@zu.de
168 EDUARD BRUCKNER
[Continuedfrom previous page: Secular variations ofthe climate}}
Represented by the Length of periods as given
variations of ice conditions by
of
Grape Frequency
Lakes Glaciers Harvest of Cold
Dates Winters
Cold 74 4) 1677-81 30
Warn 83 5) and 40
Cold 1707-146) 1710-16 30 40
Warm about 1720 45 45
Cold about 40 50--67 50 45
Warm about 60 60-86 30 35
Cold about 80 about 1800 20 25
Warm about 1800 1800-15 40
Cold about 20 15-30 45
Warm about 35 30-45 30
Cold about 50 45-75 35
Warm about 65 75-90 30
Cold about 80 35
In determining the average value of five successive variations we arrive at the
following:
For the Time Span 1020-1190 34 years
1190-1370 36
1370-1545 35
1545-1715 1) 34
1715-1890 35
The length of the period remains almost the same irrespective of the time
span selected.
Given these findings we can now calculate the probabale error for the
mean period of 34.8 years with Fechner's formula. Given the mean deviation
of±7.0 years, this error amounts to ±0.7 years. The uncertainty is ± 5.9 years
for each individual period. This figure lets us conclude that it would take 544
cold and warm periods or in other words 272 complete climate variations in
order to pinpoint the average length of a period on the basis of lustra
averages down to one month only; that would mean approximately 9500
years of observation. Yet even today we can be quite satisfied with our
4) Neusiedler See.
5) Lake Fucin.
6) Lake Zirknitz, Caspian Sea.
1) The cut-off date is 1715 according to the ice conditions of rivers, the water level variations
of lakes, and variations of glaciers; the year 1705 seems to be more precise according to
the grape harvest dates and the frequency of cold winters.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 169
result; after all, its margin of error does not amount to more than 2% of the
length of a period.
For the time being this evidence applies to central Europe only because
the observations of the grape harvest dates and the frequency of cold winters
are confined to this region. But we should not disguise the fact that indirectly
it may be of universal significance around the globe. The meteorological
observations gave evidence that the same climatic variations, which affected
central Europe in this century, occurred simultaneously in Inner Asia, North
America, Australia, etc., in short, with a few exceptions, all over the globe. It
is unlikely that this simultaneous occurrence is merely a coincidence and a
peculiar feature of the century. It is far more likely that we are dealing with a
regular pattern. Consequently the observations carried out in one country
were indicators of climatic variations all around the globe. Those variations
in the timing of the grape harvests and in the frequency of severe winters
which in central Europe could be traced back to the year 1000 are merely
one aspect of the global climate variations occurring since then with a
periodicity of 34.8 years, give or take 0.7 years.
We have reached the end of our discourse about climate variations. We
have established that their characteristics are identical with those of the
variations of temperature, air pressure, and rainfall, and finally we were able
to determine their periodicity from observations spanning nearly nine
centuries. Are these climate variations indeed so significant that they have a
practical impact? To show, that this is very much the case, will be the theme
of the next section:
• [In the following Section 4c.]
nico.stehr@zu.de
4.3 THE SIGNIFICANCE OF CLIMATIC
VARIATIONS IN THEORY AND PRACTICE*
Influence of the climatic variations on dimensions of glaciers; as well as on the
dimensions and drainage conditions of lakes, on the frequency of floods, and
on the water-level of rivers. The effects of these variations as well as those in
the duration of ice covers on traffic life. Links between climate variations and
agriculture, demonstrated in a tabulation of grape and wheat crops.
Anticipating a severe economical crisis in the dry regions of the Great Salt
Lake. Link to typhus frequency demonstrated in several tables. Influence of
rivers' run-off on the coastal water levels: Coasts of the Baltic Sea and the
French channel (table). Declaring some presumed shifts of the beach line as
caused by climatic variations (Paschen, Bouquet de la Grye). SueB goes too
far. Significance of climatic variations in determining mean values in
climatology, demonstrated at three stations. Prognoses based on climatic
variations. List of scientists who anticipated climatic variations: Hann,
Schweinfurth, Dove, Zimmermann, Plantamour, Lorenzoni, Kluge, Hagen,
Marie Davy, Jevons, 1. A. Brown, and perhaps also Fritz. First clearly defined
proof by Sonklar, but limited to the Alps. The universal significance of
climatic variations has so far not been recognized. Climate variations are
reflected in the history of views about climate change.
It is not my intention to describe in detail how important climatic variations
are for the multitude of issues pertaining to practical life and to science. I
may be permitted a brief outline however. In doing so, I shall refrain as
much as possible from generalizations and instead do my utmost to attempt a
quantification of this impact. The enormous effect on glaciers is well known.
It was, after all, their varying sizes that made us first aware of the occurrence
of climate variations. Here is one example for the enormity of these
variations. Sonklar measured the glacial area of the Hohe Tauern range
[High Tauern] based on surveys which were carried out shortly before and
immediately after the glaciers had reached maximum sizes and found it to be
422 sq. km. I repeated his measurement on the new local map of this
Austrian area which was put together in the early 1870s, i.e., at a time, when
the glaciers had already been receding for almost two decades. The resulting
area of 363 sq. km. was 14% smaller than in Sonklar's measurement. 334 As
• [Chapter 9: Die Bedeutung der Klimaschwankungenfiir Theorie undPraxis]
334 Compare this with Bruckner: Die Hohen Tauern und ihre Eisbedeckung [The Hohen
Tauern mountains and their ice cover], Zeitschrift des Deutschen und Osterreichischen
Alpenvereins, 1886. Richter (Die Gletscher der Ostalpen [The Glaciers a/the East Alps].
Handbiicher zur Deutschen Liinder- und Volkskunde, Volume 3, 1888) determined a
glacier area of 381 sq. km in accordance with the same dimensions, resulting in a decrease
of 10 percent since Sonklar's time. The difference between this percentage figure and my
171
nico.stehr@zu.de
172 EDUARD BRUCKNER
glaciers in the Eastern Alps continued to recede until the end of the 1880s, a
decrease of about 20% of the area defined by Sonklar is more than likely. In
accordance with our climatic variations the size of the glacial area in the
Hohen Tauern range changes by approximately this amount. Based on his
observations at eight glaciers of the Eastern Alps, Richter determined that
during the most recent stage of this decrease glaciers lost an average of 6.17
cubic metres (cbm) of ice per sq. m through melting. 335 For the entire region
of the Eastern Alps this amounts to 9 cubic kilometres and for the entire area
of the Alps to 25 cubic kilometres, or to the equivalent of an ice cube of
nearly 3 kilometres in length, width as well as height. In volume, this cube
would in turn be the equivalent to a 15 cm-thick layer of ice covering the
entire Alps; the average elevation of the Alps has during the last 20-30 years
decreased by 15 cm.
, There are other ways in which our climatic variations are able to
temporarily modify the geographical picture of a region extensively. In dry
regions in particular, where water is scarce to begin with, hydrological
conditions change tremendously in the course of variations iIi rainfall. A
map drawn during a dry season will often look totally different from one
produced in a wet period. Lakes disappear during dry periods, only to
reappear in wet periods, at! for instance Lake George in New South Wales,
which around 1820 and 1876 and to a lesser degree in 1850 was a
magnificent lake, 20-30 km long,336 10 kilometres wide and 5-8 meters
deep. During the intervening dry periods however the lake vanished entirely
from the face of this earth and grass grew in its basin. The neighbouring
Lakes Cowal and Bathurst dry out completely during the dry season and
reappear during the following wet season. The enormity of this fact is fully
comprehensible when we consider that at high water levels Lake George as
well as Lake Cowal cover a region comparable to the area of Ziiricher See.
Similar conditions are found at the swampy Lake Hamun in Persia, although
this lake does not disappear entirely. The Great Salt Lake, again, undergoes
tremendous changes, its surface area increasing by a full 17 percent from its
last minimum level in the 1850s to its maximum level in the 1870s, while
Lake Fucine's decreased by 19.2 percent between 1816 and 1835. The
changes which occurred in the Caspian Sea are smaller in relative terms, yet
much more considerable in the absolute. When from 1809 to 1814 until the
beginning of the 1840s, the Sea's level fell by 3 meters the overall area was
own can partially be explained by the fact that my measurements were based on a special
map I :75,000 whereas Richter used the original maps of 1:25,000.
335 Richter, op. cit., p.297.
336 In my lecture (Verhandlungen des Berliner Geographen-Tages, Berlin 1890) the length
was erroneously stated to be 18 Ian instead of 18 miles.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 173
reduced by 3 percent or 13,000 sq. kms (square kilometres).337 Climatic
changes are demonstrated in another and equally impressive way at several
Central African lakes; according to Sieger the Tsad, the Tanganyika and the
Nyassa lakes rise at times to such high levels that they continue to run off for
as long as a whole decade, then stop doing so completely in the course of a
drought. The same applies to the Abistada Lake in Afghanistan and possibly
the Goktscha Lake in Armania. These are changes that even a school-atlas
must map out.
Also affected are flowing bodies of water. Rivers and creeks dry up for a
whole decade; swamps dry out but reappear in the next wet period.
So does the Atrek river. According to G. Sievers the Turkomans claimed
that its water levels have dropped considerably over a number of years prior
to 1871 and that in the summer it does not even reach the delta area, etc. 338
During the rainy periods destructive floods are again a common occurrence,
as for instance in Australia and especially in New South Wales where
according to Jevons floods occurred as follows:
Time Span 1799-1821 1822-1841 1842-1858
Number of Floods 14 4 10
Probability 0.61 0.20 0.59
Jevons therefore emphasises the fact that time periods whith a great
number of floodings alternate with those marked by extremely low water. 339
Russel later wrongly opposed this statement and replaced it with his own
hypothesis of a 19-year period (compare with Chapter 4a). As a result of
climatic variations very dry regions suffer from extremes of this kind. But
even regions rich in rainfall and water are affected, although somewhat less
severely because the variations of the rainfall are less pronounced in these
regIOns.
Through these hydrological phenomena climatic variations have a
profound effect on the human environment. River-navigation is highly
dependent on the river's water volume which determines the river's depth.
As a result, the latter is, in accordance with the climatic variations, either
increasing or decreasing. When the five-year average of the water levels of
Seine, Donau, Rhein, Weser, Elbe, Oder and Weichsel was half a meter and
even lower during the dry seasons around 1830 and 1860 than it was in the
rainy periods around 1815, 1850, and 1880, river-navigation had to be
severely affected because a drop or rise by this amount makes a big
337 Defined with the help of the hypsographical curve.
338 Sievers, in Petermann's Mittheilungen, 1873, p. 292.
339 Jevons, in Waugh's Australian Almanach 1859, pp. 61-76.
nico.stehr@zu.de
174 EDUARD BRUCKNER
difference. Navigational problems did indeed increase during this time, and
soon a huge amount of literary works was produced dealing with the
question of the cause for this drop in the river water levels. In most instances
increased deforestation was found to be the cause. Nowadays we know
better: It is the climate variations that bring about periodic changes affecting
river traffic one way or the other.
Varying temperatures affect river traffic in another way through the
duration of the river's ice cover.340 In the middle of a cold period, e.g., the
coldest lustrum, navigation is according to the averages of several variations
and stations closed down for a period of 16 days longer in Siberia and in the
Ural, 20 in Northern Russia, 17 in South East Russia, 25 in the Baltic
provinces and 32 even in South West Russia including the Donau and the
Weichsel regions than in the middle of a warm period, e.g., warmest lustrum.
This influence is evident even in the averages of the cold and warm periods.
At Kronstadt, for example, the annual navigation period was as follows: 341
1814-21 1822-36 1837-56 1857-63 1864-83
Averages (in Days) 200.5 231.7 204.5 223.4 205.0
Number of Years
II 7 5 7
at Normal Levels 342
Number of Years
8 4 13 2 13
Below Number
Number of Favourable
Years in Percent
o 73 35 71 35
Consequently, the piers of Kronstadt and also the harbour of st.
Petersburg remained closed over an average of 3-4 weeks longer during cold
periods than during warm periods; that amounts to 1110 to 117 of the entire
navigation period. It means that during the cold period those harbours with a
more westerly location and shorter closure times receive part of St.
Petersburg's shipping traffic that they lose again during the warmer period.
Of course the length of the navigation period may also change from year to
year. Yet in that case, the following year compensates for the loss of the past
year. This is not the case with climatic variations where average values have
changed and where either condition, favourable or unfavourable, remains the
same on a multi-year average. In the cold periods only one third of all years
is favourable, meaning the navigation period is longer than normal, whereas
two thirds are unfavourable; in the warm period however two thirds to three
quarters of all years are favourable and only one third to one quarter
340 See Chapter 4b.
341 Calculated by Rykatschew: Open and Closure of the Rivers in Russia, St. Petersburg,
1886, p. 294 (in Russian).
342 208 days. The abnormal years 1823 and 1835 are not included in this average value.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 175
unfavourable. The conditions at Helsingfors are quite similar to those at
Kronstadt. In the wann lustrum of 1831-35 the pier at Helsingfors was free
of ice about one week earlier than in the cold lustra of 1836-55, as were the
piers of Arensburg and Pernau. The same phenomenon is to be found on the
American continent. Shipping on the canal of Lake Erie in the State of New
York was shut down annually as follows: 1828-33, 129 days; 1838-47, 141
days; 1848-57, 135 days. In this way climatic variations go hand in hand
with certain changes in trafficabi1ity which are by no means unimportant.
Another area where the influence of climatic variations is felt is
agriculture. This influence, however, differs considerably according to
location. In areas with an abundance of moisture where consequently wet
years produce lower crop yields, the dry periods of our climatic variations
yield richer crops. In a dry climate, however, where drought and crop failure
go hand in hand,. it is the dry period in particular in which yields are low. To
substantiate this correlation I tabulated the yields for the individual years on
the basis of data submitted by Fritz. 343
A few words about what these figures represent:
The vintages of Volnay, Aargau, Nassau, Wiirttemberg and France are in
hectolitres per hectare annually; for Prussia in buckets (of 68.7 1) per
Prussian acre (0.26 hectare); for the Domline Hochburg in Baden and for
Hessia again in hectolitres per hectare, for the state of Ohio in gallons per
acre. 344 The wheat crop is in bushels per acre. 345
In the dry and wann years around 1830 vintages are consistently higher
in Central Europe. Between 1840 and 1855 they are very low, while around
1860 they increase. Towards 1880 they start to decrease again. It is just the
reverse in Ohio; here the maximum rainfall of 1876/80 coincides with a
maximum vintage. That conditions in Ohio are reversed in comparison to
Central Europe, becomes quite evident when we look at the figures for the
wheat crops: crops were poor in the dry 1860s and rich in the moist years
around 1880.
Even in earlier years this correlation existed in Central Europe. Low
yields were produced in the years 1576-90, 1765-74, 1812-17; high yields
in 1671-78, 1698-1708, 1818-28, etc. Grain prices in England were high
pointing to, at least in general tenns, low crops around 1648, 1700, 1810 and
1855; in Central Europe around 1760, 1817; in Southern Gennany and in
Switzerland around 1544, 1587, 1710, 1795, 1817, 1847 and 1855. In
Central Europe vintages were good in the years 1470, 1534, 1636, 1678,
343 Fritz: Periodische Erscheinungen der Meteorologie und Kosmologie [Periodical
Phenomena of Meteorology and Cosmology], Intemationale wissenschaftliche Bibliothek
Volume 68, Leipzig, 1889.
344 Fritz, op.cit., p. 283, 296; only France according to Angot, op. Cit, p.33.
345 Fritz, op. cit. ,po 303.
nico.stehr@zu.de
176 EDUARD BRUCKNER
Secular Variations o[Gra2.e and Wheat Croe. Deviation.t!:.om the mean
GraEe Harvest WheatCroE
Volnay, Aargau, tlI)
o!. §
Nassau U .!!!
U !:l
:fa ::8 U = State
Wiirttemberg
~
~ ~
'"E
'" 8 g
Cl::t:
..cifoU
8
c:Q
~
'"
o2::t: Ohio
of
State of Ohio
Average 15 ? 7 28 26 70 13.3
1821-25 -6 -I 12.
26-30 6 2 8
31-35 5 3 17.
36-40 3 -2 -6
41-45 -4 3* -15.*
46-50 3 2 0
51-55 -7* -8* -2 -14 ~.1
56-60 2 -2 -6 ~.8
61-65 -1 2 -2.6*
66-70 3 5 9 -4 -1.2
71-75 0 6 -5 -18 0.4
76-80 -7* 3* -8* 14 2.3
81-85 -I 2. 7 2.0
1724, 1784, 1854; on the other hand they were low around 1440, 1485,
1605, 1695, 1765, 1810, 1875; for Southern Germany, the Eastern part of
Switzerland, the Central Rhein and the Mosel around 1482 93, 1595-1610,
1685-1700, 1755-74 and 1795-1820.346 If we compare these years with the
cold and warm or wet and dry periods which we established earlier in
connection with the frequency pattern of cold winters and the timing of the
grape harvest (Chapter 4b), they fall into the same periodic pattern, with the
exception of those years marked in italics. If we simply count the number of
exceptions as well as the number of those years which follow the established
periodicity, regardless whether they depict a single year or longer time
spans, we arrive at 38 for the periodic years, while there are only 11
exceptions. In other words, in Central and Western Europe 77 percent of
those years with particularly rich crops fall into the warm and dry periods
and the same percentage of years with poor crops falls into the moist and
cold periods, which fully supports our fmdings about the past 60 years.
In dry regions as well as in the tropics in general, aridity is detrimental to
crop yields. In the 1860s Mauritius for example suffered tremendously from
droughts that were attributed to the rampant practice of deforestation, yet
disappeared when rainfall increased after 1865 and especially after 1870.347
Siberia, too, suffered severe droughts around 1860 and as a result had a
346 Fritz, op. cit., p. 265, 387.
347 Compare Koppen in the Annalen der Hydrographie, 1887, p. 280.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 177
number of crop failures. According to Jevon,348 in New South Wales the
years 1799-1821, in general a rainy period, show only three prominent
droughts, however the dry period of 1822--41 shows nine and the wet years
of 1842-57 again only three. As reported by Darwin/49 in the years 1827-30
terrible droughts occurred in the provinces of Buenos Aires and Santa Fe at
the Parana river killing millions of animals. In 1791 and 1792 East India, the
West Indies, and the Cape Verde Islands suffered droughts simultaneously.
Both years fall into the middle of two dry periods. Of the droughts in this
century, as documented by Fritz350 for East India, New South Wales, North
America and Mauritius, 60% occurred in dry periods, although there have
been three rainy periods and only two dry periods since 1800.
Climatic variations are influencing the fertility of Egypt's agricultural
land in a very particular way. Agriculture's total dependence on the waters of
the Nile is well-known. A rise of the Nile's level by only a few centimetres
turns vast, normally infertile regions into moist, fertile and productive areas.
According to our table: in 1846-50 the Nile levels rose on an annual
average 0.86 m above the level of 1831-35 and 0.38 m above the one of
1856-60, as well as 0.60 m above the 1856--60 level in 1871-75. Such
fluctuations are of eminent importance to the productivity of these
agricultural regions.
The variations of the climate should turn out to be downright disastrous
for the future of the dry regions of Central North America located around the
Great Salt Lake. From the early 1860s to the middle 1870s the level of the
Great Salt Lake rose by 3 m; its tributaries filled with water which was then
used for the irrigation of newly cultivated fields and grasslands. 351 As we
have learned earlier, general opinion agrees that the expansion of cultivated
land in formerly arid regions has increased rainfall considerably.
In contrast I would like to point out that the improvement of the climate
takes place in exactly the same time period in which as a result of climatic
variations rainfall increased over more or less all of the global land masses,
but particularly over the continental regions. That, on the other hand, similar
changes of the climate, to the worse or the better, occurred in earlier time
periods as well in America, was established for the current century from
observations of the rainfall and the river levels in the neighbouring
Mississippi delta, and for far back into the previous century on the basis of
temperature observations. These climate variations are identical with those
348 Jevon, op. cit.
349 Darwin's Scientific Explorations, etc., German by Dieffenbach, 1844, Volume I, pp. 151,
153.
350 Fritz, op.cit., p. 331 .
• [in a Chapter not contained in this excerpt]
351 Compare Gilbert, in Powell: Report on arid regions, Washington, 1879, p. 55 ff.
nico.stehr@zu.de
178 EDUARD BRUCKNER
in Europe which could be traced back to the year 1000. The large number of
established variations compels us to believe that they will continue to occur
in the future. All facts considered it appears most likely that the period of
climatic improvement which can be observed at the Great Salt Lake since
the early 1860s, will be followed by a period of climatic deterioration. First
indications of this development can be detected in the dryer climate in recent
years resulting in a lower level of the Great Salt Lake: in 1888 the lake had
already reached the previous low of 1864.352 Should this assumption become
fact, these areas will inevitably suffer a severe economic crisis very soon,
because the land which used to be cultivatable in the years 1870-80 would
quite soon be unproductive as a result of the drought.353 Then here, too,
would become apparent what we have seen happening in Egypt and Siberia:
not only will yields vary in size along with these climate changes but also
the actual amount of cultivatable land.
Climatic variations again are not without impact on health conditions. I
have, however, examined this correlation in one direction only by
endeavouring to establish the link between climatic variations and the
occurrences of typhus. Weather conditions do not directly affect the
frequency of typhus cases but rather by way of the ground water. Even
though Pettenkofer's groundwater theory may still have many critics, hardly
anyone who has familiarised himself with the evidence will be able to ignore
it completely.
The groundwater level falls and rises parallel with climatic variations; it
was high around 1850 and around 1880 during the moist and cool periods
and low around 1830-1860 during the dry and warm ones.
Secular Variations of the Groundwater (m)
1856-60 61--65 66-70 71-75 76-80 81-85
Munchen 0.09* 0.27 0.42 0.27 0.75 0.35
Salzburg 2.97* 3.03 2.99 3.13 3.04
Penck was the first to suspect that consequently epidemic outbreaks, too,
might be influenced by climate variations. 354 I was able to provide statistical
proof of this correlation first for Hamburg and then for other regions as
well. 355 Unfortunately, morbidity rates were unavailable, I had to rely on
352 According to the hand-written curve by G.K. Gilbert provided to me by Dr. R. Sieger.
353 I mentioned this conclusion in a public lecture entitled Is our climate changing?, held
March 31, 1888 in the auditorium of the University of Dorpat (compare with article in No.
68 of Neue D6rptschen Zeitung).
354 Penck in the Milnchner Allgemeinen Zeitung, end of 1887, in an essay about Soyka, Der
Boden [The Soil].
355 Compare for Hamburg, Bruckner: Grundwaser und Typhus [Groundwater and Typhus,
Chapter I] Verhandlungen der geographischen Gesellschaft in Hamburg, 1887-88,
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 179
mortality rates only. These however may give a distorted view of the actual
circumstances because they may be affected by medical changes. Even the
link between morbidity rates and groundwater levels is of course,
particularly in recent decades, no longer a true functional correlation because
of the rapid progress in improvements to urban sanitation systems. As a
consequence, the figures for Basel are the only ones that show a parallel
development of typhus mortality and climate change while at other locations
the mortality rate has generally been declining since it was first recorded.
I will begin with the data for Basel, which have been smoothed by
forming five-year averages.
In the first half of the twenties the mortality rate is low (below 1 per
10,000 inhabitants), from 1828 to 1849 however it is high with a maximum
°
in 1832 and a break in 1836, low again from 1850 to 1853 (minimum 1853),
quite high in 1854-68 (maximum 1864), and once more low from 1869 to
1883 (minimum 1877). If we ignore the first four years the records of which
seem unreliable/56 and determine the mortality rate for those periods which
we identified as wet and dry, we get the following results: dry period 1825 to
1840, 11.4; rainy period 1841-55, 12.0; dry period 1856-70, 21.7; rainy
period 1871-83, 6.2. The only significant deviation occurs in the first half of
the forties, in which the mortality is high despite the fact that these are years
with a lot of rainfall.
Annual Typhus Mortalities at Basel per 10,000 Residents.
Five-vear
- A
-----._--- --
1824 2.4 1839 14.6 1854 11.9 1869 8.8
25 3.2 40 15.1 55 12.5 70 6.3
26 3.0 41 17.6 56 18.1 71 5.4
27 3.1 42 17.2 57 20.0 72 5.5
28 11.4 43 14.6 58 23.2 73 5.8
29 12.6 44 13.8 59 23.2 74 5.4
30 13.6 45 13.8 60 24.7 75 6.1
31 16.4 46 12.1 61 19.6 76 6.1
32 17.8 47 11.6 62 17.8 77 5.3*
33 10 .0 48 12.3 63 24.8 78 6.1
34 10.1 49 11.1 64 30.1 79 8.3
35 10.5 50 9.6 65 29.9 80 7.2
36 8.6 51 9.4 66 29.4 81 6.7
37 10.1 52 9.3 67 28.2 82 7.7
38 12.9 53 9.1* 68 17.1 83 7.0
Volume 3. The remaining rows have been calculated according to those in Reincke: Der
Typhus in Hamburg [The Typhus o/Hamburg], Hamburg, 1890, p. 68.
356 Small values like this do not occur in later years.
nico.stehr@zu.de
180 EDUARD BRUCKNER
That such variations have not taken place in other cities, is demonstrated by
the next table, which contains Basel for comparison as well
It is obvious that, in general, from the year 1860 on, typhus mortality
rates are falling; however, this decline has to be seen mainly as a
consequence of the steady improvements in the sanitation systems, since all
locations show this tendency towards lower figures as the years progress. In
my opinion, this improvement since 1860 can at least in part be attributed to
rising groundwater levels. Indicative for· this correlation is, to me, the
mortality decline during the relatively dry lustrum of 1871-75 in Miinchen,
Frankfurt a.M., Berlin, and Chemnitz, as well as its rise during the dry
lustrum of 1856-60 in Frankfurt a.M., Chemnitz, and Basel as compared to
the previous wet lustrum. The increase of the rate from 1820 to 1836 in
Hamburg conforms again to the then prevailing dry period.
If we approach the issue in a slightly different way and, instead of
concentrating on the number of typhus mortalities within a given lustrum,
determine the difference from the previous lustrum, the influence of the
climatic variations is clearly in evidence. This step has been taken in the next
table. The figures show by how much in a given lustrum the typhus mortality
per 10,000 inhabitants has increased (+) or decreased (-) in comparison to
the previous lustrum.
With a few exceptions, it emerges that the dry periods around 1830 and
1860 show deterioration or at least a complete halt in the general
improvement of the typhus mortality rate, whereas the wet periods show a
rapid improvement. The table's last column displaying the averages of all
other columns shows these variations in the improvement very clearly. The
maxima fall exactly into the wet lustra of 1846-50 and 1876-80, the minima
into the dry lustra of 1826-30 and 1856--60. These findings underline the
probability that climatic variations have a significant impact on the mortality
rates of typhus.
Annual Typhus Mortalities Expressed in Lustra Averages per 10,000 Inhabitants.
Hamburg MOOchen Wiirzburg Augsburg FrankfurtlM.
1821-25 9.7
26-30 13.6
31-35 14.2
36-40 15.1
41-45 15.9
46-50 12.8 13.8
51-55 10.6 14.4 8.5
56-60 8.4 24.0 11.6 8.8
61-65 7.7 18.7 13.2 11.6 5.0
66-70 7.2 11.9 15.7 11.9 5.7
71-75 5.6 15.6 6.9 6.6 6.8
76-80 3.1 7.7 3.1 5.1 2.1
81-85 3.0* 1.7* 1.8* 1.5* 1.2*
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 181
Badisches
Berlin Ann;y-co!,Qs Baden Chemnitz357 Basel
1821-25
26-30 13 11.2
31-35 34 10.0
36-40 50 12.9
41-45 58 14.6
46-50 59 12.3
51-55 27 13.6 6.2 9.1
56-60 10.0 16 10.5 8.0 23.2
61-65 9.6 13* 8,5 5.6 24.8
66-70 8.6 14. 7.5 3.2 17.1
71-75 10.1 6.9 4.5 5.8*
76-80 4.6 3.8* 1.6* 6.1
81-85 2.6* 2.5 7.0.
As paradox as this may at first sound to some, it is undeniable that the
water levels of the ocean are affected by climatic variations. 358 In his
standard work about the currents of the Northern Sea, Mohn has
demonstrated how many interdependent factors are involved in determining
the sea level at any given moment in time. There is the air pressure, which in
its irregular distribution alters the sea level, or the ocean's salt content that
near the coasts is greatly affected by any fresh-water influx. There is also the
wind's impact, which should neither be under- nor overestimated, and finally
there is the temperature which in combination with the salt content
determines the water's density.
The distorting effects that all these factors combined produce, are quite
severe and reach one meter on Mohn's chart. Of this amount the factors'
proportion is as follows: density of the sea water 0.7 m at maximum level,
wind 0.8 m, air pressure a mere -0.07 m, i.e., the air pressure cancels out a
small portion of the other factors' distorting effect. The sea level is lowest
halfway between Iceland and Spitzbergen as well as between Greenland and
Norway, and highest close to Greenland and at the estuary of the Baltic Sea.
358 Mohn: Nordhavets Dybter, Temperatur og Strominger [The Nordic Sea, Temperatures and
Currents], Volume 8 of the essays of Norske Nordhavets-Expedition, 1876-78.
Preliminary infonnation which however deviates quite substantially from the final version
published as supplement No. 79 to Petermann's Mittheilungen, Gotha, 1885. [The
Norwegian tenn "Nordhav" is today uncommon. It seems that Mohn's suggestion to name
the sea between Norway, Faroe Islands, Iceland, Jan Mayen and Spitsbergen "Norwegian
Sea" was accepted. (Mohn: Die Norwegische Nordmeer-Expedition [The Norwegain
Nordic Sea expedition] (Petennann's Geogr. Mittheilungen, 1878).]
nico.stehr@zu.de
182 EDUARD BRUCKNER
Change of Typhus Mortalities from one Lustrum to the Next
Hamburg Miinchen Wiirzburg Augsburg Frankfurt a.M.
1826-30 +3.9
31-35 +0.6
36-40 +0.9
41-45 +0.8
46-50 -3.1
51-55 -2.2 +0.6
56-60 -2.2 -2.8 0.3
61-65 -0.7 -5.3 +1.6 -3.8
66-70 -0.5 -6.8 +2.5 +0.3 +0.7
71-75 -1.6 -3.7 -8.8* -5.3* +1.1
76-80 -2.5* -7.9* -3.7 -1.5 -4.7*
81-85 -0.1 -6.0 -1.3 -3.6 -0.9
Baden Chemnitz
359
Berlin Arml::-Co~s Baden Basel Average
1826-30 +3.9
31-35 +21 -1.2 -0.3
36-40 +16 +2.9 +1.9
41-45 +8 +1.7 +1.2
46-50 +1 -2.3 -2.7*
51-55 -32* -3.2* -1.6
56-60 -11 -3.1 +1.8 +14.1 +1.4
61-65 -0.4 -3 -2.0 -2.4 +1.6 -1.4
66-70 -1.0 +1 -1.0 -2.4 -7.7 -1.5
71-75 +2.1 -0.6 +1.3 -11.3* -2.9
76-80 -5.5* -3.1* -2.9* +0.3 -3.6*
81-85 -2.0 +0.9 +0.9 -1.5
As we partly know already, and can partly a priori assume, all these
factors reflect the climatic variations. To show this variability for the two
factors wind and temperature, is however not possible today because of the
lack of adequate observations. Having learnt earlier about variations of the
air pressure, those of the salt content can be demonstrated for the Baltic Sea.
Its salt content was high from 1869 to 1873, decreasing after that. In the
middle of the rainy period, that is 1878-81, off Rugen it was two per
thousand lower than around 1870, and in the vicinity of the Little Belt even
five to six per thousand. These data appear to point to a decrease in salt
content during wet periods and a concentration in dry periods. However
these figures must be treated with caution because of the short observation
time, the more so since according to Karsten the salt content in the Baltic
Sea is mainly influenced by storm tides from the West. Be that as it may, we
are in a position to identify fluctuations of the ocean level at different
measuring stations, displaying the same periodicity as the climatic
variations, as I have stated in 1887 in my lecture at the assembly of the
359 Hospital admittance in percentages.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 183
German Meteorological Society.36O Of the series of sea level observations
that I used at that time, I will repeat the data of Swinemiinde at this point,
raw as well as smoothed, and will compare theJ;Il with the water level
observations at the Oder River. In addition, I am including the lustra
averages of the water level observations at Brest, Cherbourg and Le Havre,
which were compiled by Bouquet de la Grye. 361
Secular Fluctuations of the Sea Levels Adjusted b~ Standard Deviations
Swinemiinde {mm} Oder, Neuglitzen {m}
Raw Smoothed Raw Smoothed
1826-30 -7* -3* 2.1 2.0
31-35 5 -I 1.7 1.9*
36--40 -5 0 2.2 2.0
41-45 6 2 2.0 2.2
46-50 0 6 2.5 2.4
51-55 19 6 2.7 2.4
45--{)0 -14 -10 1.9 2.1
61--{)5 -33* -10 1.8* 1.9*
66-70 41 7 2.1 2.0
71-75 -12 -16 2.0 2.1
76-80 47 21 2.4 2.3
81-85
Raw
LeHavre Chrerbourg Brest
mm mm mm
1826-30
31-35
36-40
41-45
46-50
51-55 10
45--{)0
61--{)5} -27* I -17
66-70 19 -10
71-75 5 -13* -23*
76-80 52 31 --{)
81-85 -17 21 7
Around 1830 and 1860, at Cherbourg and Brest a little later, water levels
are low, yet around 1850 and 1880 they are high without exception.
That this conformity in the two periodic patterns of sea level and climate
should be merely accidental is out of the question, especially since the rest of
360 Compare Annalen der Hydrographie, 1888, February issue.
361 Bouquet de la Grye, in the Comptes Rendus of Paris 1888, II, p. 813.
nico.stehr@zu.de
184 EDUARD BRUCKNER
the stations in the Baltic Sea show a similar pattern. Consequently I have
held the opinion that the fluctuations of the Baltic Sea as well as of the Black
Sea are the result of a changing influx of river water. They indicate that there
is sometimes more, sometimes less water in these seas which are relatively
detached from the rest of the ocean .. 362 Inquiries, which I carried out later,
convinced me however that salt content, too, plays an important role. 363 It
varies indeed from lustrum to lustrum according to the amount of river water
dispersed into the Baltic Sea in the vicinity of the observation station, and
this correlation is more pronounced the closer the station's vicinity to the
delta of a big river.
Lorenz had shown364 that the sea level at a river's estuary forms a very flat
cone with its peak pointing into the centre of the estuary. The cause of this
phenomenon lies in the fact that in this spot the salt content of the water
increases into every direction. The stronger the fresh-water influx, the higher
the cone's peak, and vice versa. That is indeed the case along the coasts of
the Baltic Sea, as a comparison between sea level observations of stations
located near a river's estuary and those further away will confirm. Secular
variations of the sea level have different amplitudes, which tend to be most
pronounced near a river estuary. Of special interest is the fact, which I will
merely allude to at this point, that along with the climate variations during
wet periods the drop of the sea level is steeper the further it is away from the
fresh-water entrances, and flatter along with the varying climate during dry
periods, because this fact helps us to understand the variations in the sea
level near the coasts as we see them at Brest, Cherbourg and Le Havre. And
indeed the fluctuations at Le Havre at the Seine's estuary are much larger
than in Cherbourg and Brest, both far from any river estuaries, namely, 79,
44, and 33 mm.
It follows from all this that the sea level along coastlines and in relatively
detached parts of the ocean rises and falls in accordance with climatic
variations.
This knowledge sheds some light on the significance of several attempts
to draw conclusions from sea level observations as to the rising and falling
of the coastline. When Paschen on the basis of his 1849-66 observations at
Wismar took a 1520-to-l bet that the coastline would rise, it is because his
observations were carried out at a time when the rainfall variability showed a
downward trend. In fact the observations of the following 15 years, when
rainfall increased again, have reversed his findings. Of exactly the same
value is Bouquet de la Grye's finding, that, judging from the sea level
observations mentioned above in condensed form, the ground near Le Havre
362 Bruckner, op. cit.
363 Bruckner, in Naturforscher, Tiibingen 1887, pp. 291-293.
364 Lorenz, Sitzungsberichte der Wieder Akademie, 1863,2, p. 612.
nico.stehr@zu.de
CLIMA TE CHANGE SINCE 1700 185
gives way by 2 mm annually, at Cherbourg by 1 mm, and at Brest not at all.
If as far as Le Havre and Cherbourg are concerned Bouquet de 1a Grye had
been looking at observations from the early fifties as he did for Brest, he
most likely would have come up with the same result for these stations as he
did for Brest.
Although I am thus inclined to see some of the alleged changes of the
coastline as the result of climatic variations, I would not under any
circumstances go as far as SueB does, who attributes shifts of several feet as
they have been observed near water-marks for more than 100 years in
Finland and Sweden, to climatic variations. 365
This discourse leads automatically to the question of the significance of
climatic variations and their 35-year periodicity for determining
climatological mean values, the so-called standard values. That they must
play a role goes without saying. As soon as periodical variations have been
established, a mean value will only then approach the standard value if the
time periods of opposite characteristics are entered as factors of equal
weight, i.e., as quantitatively equal factors, or in other words, if the number
of observed years is a multiple of the periodicity of those variations.
Otherwise the mean will deviate sometimes more sometimes less from the
standard value. As a consequence, 35-, 70-, 105-year, etc., averages are
much closer to the standard value than those from time periods which are not
a multiple of the 35-year period. The empirical approach confirms this. For
the two observation series Praha and Madras, both of which display an
extremely regular pattern of rainfall variations, I determined several multi-
year averages. Instead of using absolute amounts I relied on the lustra
averages expressed in percentage of the 1851-80 averages as shown earlie·.
The average value nearest to the standard is in this case the mean of two
complete variations, that is at Madras the 1821-85 mean of 102.0% and at
Praha the 1816-85 mean of 101.4%. The multi-year averages, dated back
from 1885, are as follows:
No. of Years 30 35 40 45 50 55 60 65 70 75 80
Praha 100 + 1.7 2.0 3.8 4.0 4.2 2.5 2.0 1.4 1.4 1.9 1.9
Madras 100 + 2.2 2.4 4.9 4.8 5.1 2.4 2.8 2.0 3.3 2.6
It is evident that indeed the 35-year mean is closer to the standard value
than any of the other multi-year means. Its standard deviation is only 0.4%
and 0.6% respectively, while the 50-year mean deviates by 2.8% and 3.1%
respectively, i.e., by more than five times that amount. The great advantage
of the 35-year mean is obvious. It is only surpassed by the 30-year mean
365 SueB: Antlitz der Erde [Face a/the Earth], II. Wien, Praha, Leipzig, 1888.
• [in a chapter not contained in this anthology.]
nico.stehr@zu.de
186 EDUARD BRUCKNER
which deviates from the standard by only 0.3% and 0.2% respectively. The
reason for this lies in the fact that the last variation of the rainfall occurred
over an average of 30 years instead of 34.8.
It is worth noting that all of the multi-year averages cited above are
higher than the standard value, none is smaller, that is they are not to both
sides of the standard value. It could not be otherwise. In going back in time
from the year 1885 we are starting with the end of the last rainy period. This
last wet period and the previous dry period are both included in the 30-
respectively 35-year mean two wet periods and a dry one in the 50-year
mean, etc. If we start off with wet lustra the number of dry periods can never
be higher than that of the wet periods. It can at most equal that number. If we
had started off with a dry period the multi-year averages would have been
below the standard mean with the exception of the 35- or 70-year averages
which are nearly equal to the standard mean.
The two stations, which we used in this example, are located in regions
where the variations in rainfall are not very pronounced, yet we see a
substantial influence on the averages. Naturally, this influence is even
stronger where variations are more pronounced as for example in Western
Siberia. The observations at the West Siberian station Barnaul are
unfortunately limited to one periodic variation, therefore we can demonstrate
this influence only under the assumption that the variation between 1846 and
1880 represents a standard variation. 366 If we start with the year 1846 and
assume that the variations continue to occur with the same regularity as in
the years 1846-1880, each 35-year mean (101.3) represents the "standard
value". The same applies to the 70-year, 105-year mean, etc., as long as the
number of years is a multiple of 35. Should we, however, want to find this
standard value which is based on 35.x + w years of observation at not more
than 1%, it would possibly require 440 years of observation.
Consequently there can be no doubt that climatic variations are not to be
ignored in determining climatological averages that are supposed to serve as
standard values, because averages that do not include a full cycle of
variations often vary substantially from the elusive standard value. It is a
very fortunate coincidence that Hann suggested the time span of 1851-80 as
the standard period for determining climatological mean values and strictly
adhered to its application, because it covers almost exactly one complete
climate variation ..
In view of the great and far reaching significance of climatic variations
for so many areas of daily life, one is compelled to ask whether it should not
be possible on the basis of what is known about these variations to put up
prognoses for the future. In fact there can be no doubt whatsoever that these
366 The mean values of the lustra between 1846 and 1880 are as follows Ill, 101,87,66,76,
114, and 154%.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 187
periodic variations, of which we identified not less than 25 starting with the
year 1000, will continue to occur in a similar pattern in the future. With
equal certainty we can predict even more. There was every indication that a
maximum of rainfall and minimum temperatures would occur around the
year 1880; consequently we have to assume that we are presently
approaching a minimum of rainfall and maximum temperatures, in other
words a warm dry period. But this is all we can predict with certainty. When
the minimum rainfall is to be expected, can only be estimated with an error
margin of ± 6 years, which may in all likelihood be the error margin of one
single periodic variation: the minimum should occur approximately at the
turn of the century. All it means, though, is that around this time the dry and
the warm years will be more frequent than wet and cool ones. A prognosis of
this kind has no impact on Europe where rainfall variations-and these
matter most for practical purposes-are moderate. It seems to be of practical
importance only for continental regions where rainfall variations are more
pronounced: for Siberia, Australia and especially Central North America. No
doubt these regions are presently approaching a time of crop failures which
according to the latest news about the harvests in the Western States of
North America may already have begun. The approaching dry period will
most likely destroy thousands if not hundreds of thousands of livelihoods.
One may wonder at first glance why they have escaped scientific scrutiny
until now. However, there has been speculation about them. Sporadically,
most often in connection with unusual occurrences involving rivers and
lakes, the opinion is raised in publications, that the rainfall at certain
locations in particular, will for a while deviate to one side of the standard
amount and then to the other. Hann, for instance, emphasises in 1869 in a
report about various essays dealing with the increasingly dryer climate:
"Surely the effects of alternating periods of aridity and heavy rainfall are
given too little consideration."367 To what extent a similar statement by
Schweinfurth is to the point, is difficult to decide. 368
Analogous conclusions were drawn occasionally on the basis of long-
term observation series. Dove, who, by the way, opposed the idea of
meteorological cycles, recognised in 1838 quite correctly that the years
1808-24 were colder than those from 1797-1804 and 1820-30. 369
Zimmermann makes similar remarks in 1856.370 Plantamour identified multi-
year periods alternating between relatively warm and relatively cool
temperatures for Geneve. According to his findings the years 1826-34 were
mostly too warm, nearly all of the years 1835-60 too cold, and finally the
367 Hann, in Zeitschrift der Osterreichischen Gesellschaft flir Meteorologie, 1869, p. 18.
368 Schweinfurth in his preamble to Baedler igypten [Egypt] I. Th. 1877, p. 79.
369 Dove, Verhandlungen der Berliner Akademie1838, p. 345 f.
370 Zimmermann, Poggendorffs Annalen, Volume 68, 1856, p. 318.
nico.stehr@zu.de
188 EDUARD BRUCKNER
years 1861-1875, too warm again. In 1873 Lorenzoni arranged the rain
observations of the Padua station according to decades and found the
following periodicities: 371 dry periods 1733-46, 1784-95, 1812-44,
1856-71; and wet periods 1747-77, 1795-1811, 1845-55.
Both, Klug (1874) and Hagen discovered that German rivers were at
nearly the same or a lower level in the years 1817-35 and 1855-73
compared to 1836-54.372 On the other hand, the periodic time spans of high
and low water levels identified by Marie Davy for the Seine conform only in
part with our climate variations. 373
Such unusual variations were even identified for countries outside of
Europe. In 1859, Jevon, for instance, observes that in New South Wales
floods were rare and droughts unusually frequent in the time periods prior to
1798 and from 1822-41, whereas in the time spans 1799-1821 as well as
from 1842 up to the year of his writing, floods were frequent and droughts
rare.374 In 1877 John Allan Brown demonstrates that at Madras as well as at
Trevandrum in India the years 1818-27 and 1843-52 had excessive rainfall,
and the years 1828-37 and 1860--69 were remarkably dry. He concludes that
it would be of interest to find out if such variations also occur at stations
located further north. 375
Finally, at the beginning of the 1880s, Fritz compiled a number of data
from all over the world in order to demonstrate the variability of the volume
of the mainland's waters. 376 He concentrates on establishing a link between
these varying water levels and sunspots, and, consequently, fails to find out
about our climate variations. Whether he should be included in this chapter
is therefore debatable. A more likely candidate is Reis who in 1883 held the
view that the "periodic reoccurrence of water shortage and drought is linked
to sunspot activity" and who determined a primary periodicity of 110-122
years with a secondary one of 56 years. But only a few of the wet and dry
periods in his differentiation coincide with the ones we have identified,
others not at all. 377 Reis was completely unaware of our variations.
Yet none of these scientists has recognized the variations as precisely and
was as determined to verify them for a certain area on the basis of
371 Lorenzoni, Zeitschrift fUr Meteorologie, 1874, p. 188.
372 Kluge, Zeitschrift fUr Bauwesen, 1874, p. 507. Hagen, Verhandlungen der Berliner
Akademie, 1880.
373 Marie Davy, Zeitschrift fUr Meteorologie, 1874, p. 146.
374 Jevons, in Waugh's Almanach, 1859, p. 79.
375 John Allan Brown, in Nature, Volume 16, 1877, p. 333.
376 Fritz, in Petermann's Mittheilungen, 1880.
377 Reis: Periodische Wiederholungen [Periodic Repetition], etc., Leipzig, 1883. Contrary to
Reis: Lehrbuch der Physik [Textbook of Physics], Issue 7, p. 847, I would like to point out
that the fluctuations which I had identified have nothing in common with those mentioned
by Reis ifhe implies a 20-year period; his findings directly contradict facts.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 189
meteorological observations as von Sonklar did in 1858. Not until a quarter
of a century later did the works of Forel, Richter and Lang follow who again
concentrated on one particular area, the Alps.378 At no time did they mention
the possibility that these variations might occur on a far larger scale as Heim
assumes to a certain extent with regard to glacial variations. In addition, the
universal aspect of this phenomenon, its simultaneous occurrence and global
impact could not have been clearly identified before a large number of
meteorological stations had actually experienced and registered the dry
period in the 1860s and the wet period around 1880.
Our climate variations are also instrumental in solving a psychological
problem which we pointed out at the beginning of our study.
We described the clash of opinions about the issue of climate change,
some predicting a dryer or warmer climate, others claiming that it will be
moister or colder, and finally a third group that it will not change at all. The
identification of climatic variations resolves at least some of these
contradictions. In fact, in our opinion, climate does change in one direction
for a certain time and then again in the other. Depending on whether the
observations of one time period or the other are consulted, the resulting
conclusions will have to be contradictory. That this is indeed the case is
apparent from the small table below. It shows the number of essays
advocating a change of climate (wetter or dryer) for the current century in
certain parts of the world according to the time spans of their publication.
The account is based on the literature described in the first chapter including
the studies about the increase and decrease of river levels, yet excluding
those authors who advocate a kind of geological climate change on the basis
of many centuries of observations, such as Whitney, Fischer, etc. The
numbers of authors, which assessed the different time spans as wet or dry
was:
The Climate Turns
Wet D!):
1790-1805 0 1
1805-1825
1825-1845 6
1845-1860 1 2
1860-1875 0 15
1875-1888 13 7
Even if this categorisation can of course in no. way claim to be complete
because our bibliographical account was incomplete and I was not even able
to find the publication years for all of the cited works, the overall picture is
378 Heim: Gletscherkunde [Facts o/Glaciers], Stuttgart, 1885, p. 520.
nico.stehr@zu.de
190 EDUARD BRUCKNER
in essence correct. 379 It is quite evident that during each dry period and
particularly at the end of it and at the beginning of the succeeding wet period
many voices were raised assessing a tendency towards a dryer climate,
which grow silent again in the course and at the end of the wet period. They
are replaced by authors arguing for a tendency towards a wetter climate.
Four of the advocates of a dry climate fall into the lustrum of 1836/40 and
eight into the lustrum 1866170, i.e., into lustra immediately following the
height of the dry period. Climatic variations are clearly reflected in the views
on the direction of the impending climate change and their timing.
These changes in the climate and in the rivers' water levels are
predominantly attributed to changes in the earth's flora. Almost all of the
studies confirming the link between deforestation and a decrease in rainfall
are produced during dry periods, and those attributing an increase in rainfall
to reforestation during wet periods. Blanford's findings belong into this
category380 because his observations concluding an increase in rainfall due to
reforestation fall exactly into a time period when rainfall in general increased
over global land masses. Furthermore even, paralleling the climatic
variations, views about the effects of forestation are changing completely. In
the 1830s as well as in the 1860s and at the beginning of the 1870s it is
common opinion that deforestation reduces the water levels in rivers, in the
1850s however the view prevails that it is deforestation which increases the
water level. 381 The most drastic change of opinion took place in Australia. As
common as it was twenty years ago at the end of the last dry period to blame
deforestation for the increasing drought, as common was the view in the
1880s, that deforestation in particular was to blame for Australia's wet
climate. 382 "Protect the Forest" used to be the slogan; "Down with the
Forest" it is today.
In this way, climate variations guide us through at least part of the
labyrinth of hypotheses and opinions about climate change that we described
at the beginning of our inquiry. They teach us at the same time that those
diverse opinions are actually not without merit in so far as they tell us about
certain stages in the course of climatic variations. The error lies however in
the fact that the conclusions drawn correctly from these observations were
not viewed in their restrictive application to a certain time period but were
extrapolated into the past and into the future. In addition, all of the
explanatory attempts must be judged unsuccessful; the forests' influence on
rainfall, in particular, has not at all been established.
379 because some of them were not accessible to me
380 See Chapter 4a.
381 According to Schmid, op.cit., p. 25.
382 See Chapter 5a.
nico.stehr@zu.de
CLIMATE CHANGE SINCE 1700 191
We have come to the conclusion that in the course of the past nine
centuries our globe has experienced climatic variations with a periodicity of
about 35 years and an amplitude which we tried to establish for different
regions of the globe. We saw their effect on all the hydrological phenomena,
the glaciers, the lakes with no outlet, and also the rivers and river-lakes.
Through geology we have learnt about the occurrence of climatic variations
in the distant past. They were quite similar in character but differed in
frequency, perodicity and impact on organic and inorganic forms surpassing
by far the climatic variations of historic times. These variations are manifest
in the changes between ice ages and inter-glacial ages, the nature of which
remains partly in the dark. Our findings about the climate variations of past
centuries may perhaps be able to contribute to the elucidation of that
darkness.
nico.stehr@zu.de
Chapter 5
About the Influence of Snow Cover on the Climate of
the Alps *
Of all the Alps' characteristics, none appeals more to the people of the plains
than the everlasting snow covering the highest mountain peaks; nothing attracts
more than the distant glowing glaciers. The snow mass lasting throughout the
summer appears like a strange world to them because they are used to seeing
snow as a winter's companion, only appearing in the fall and disappearing in
the spring. And indeed, those glaciers are proof of an unfamiliar climate;
mountain peaks rise up to this climate zone which exists everywhere at
relatively low elevations above the inhabitants of valleys and plains and only
reaches down to the lowlands many thousands of kilometres further north in the
polar region. The snow masses of the Alps' mountain peaks most vividly
demonstrate the fact that the higher the elevation above sea level, the lower the
temperature and the colder the climatic conditions. Of course, forms of organic
life, the mountains' flora in particular, are further proof of this. But the ever so
gradual change of the flora is not as evident as the contrast between the green
vegetation in the lowland and the dazzling white snow fields of the towering
mountains, which in its overwhelming splendour fascinates us again and again.
The part of the Alps' surface that is perman~ntly buried under ice and snow
is not very large in comparison to the total range ofthe mountains. Ed. Richter
has determined the glacial area of the Eastern Alps to be 1,462 sq. km, drawing
an imaginary line reaching east of the Rhine across Mount Sp/iigen to Lake
Como. This is only 1_% of the entire area of the Eastern Alps. In relation to the
territory of Switzerland, however, the portion of the glacial area is quite large.
The total glacial area in Switzerland is 1,840 sq. km. These 1,840 sq. km make
up 4 112% of all of Switzerland and 7 1/2% of the Swiss Alps. Exact figures for
the French and Italian Alps unfortunately do not yet exist. But on the whole it
may be assumed that approximately 2 114% of the Alps' area are covered by
glaciers .
• Uber den Einjluj3 der Schneedecke auf das Klima der A/pen, Zeitschrift des Deutschen und
Osterreichischen A1penvereins, 1893.
193
nico.stehr@zu.de
194 EDUARD BRUCKNER
Given the small size of these areas, one would assume the influence on the
climate of these regions is not substantial. But it is by all means noticeable.
When we approach a glacier at the end of a valley in midsummer, we can feel a
cold wind coming from the direction of the glacier. The air stream is cooled off
by coming into contact with the ice. It then flows down the valley, often retain-
ing its low temperature over stretches of several kilometres. This flow of air
can at times become quite turbulent. Stiibel and Reiss describe such winds in
the South American Andes, caused by the difference in the temperature of air
above the glacial snow and above the snow-free valleys. These winds are so
vehement that they may overthrow horse and rider, making travelling on those
days simply impossible. Such is the force with which the cooled air funnels
down from the glacial surfaces.
If, as in the above example for the Andes, the eternal ice fields of the Alps
exert some influence on the climate of nearby valleys, it can only have a fairly
limited impact due to their small size. The snow cover changes in size through-
out the year, reaching down from the peak during the cold season and eventu-
ally covering the entire mountain range down to its base. As it expands, its
influence on the climate increases. This influence is greatest during the winter,
not in summer. Therefore, we will mainly limit our investigation to the winter
season.
In order to study the influence of the snow cover on the climate of the Alps,
it is first of all essential to have a thorough knowledge of the dimension of the
snow cover. At present there is not much research to fall back on. Few observa-
tions of the various seasonal expansions of the snow cover are available. Only
two series of observations do exist which have been carried out over a number
of years and involve the altitude levels of the bottom snow line in the moun-
tains and their monthly change. One took place at Mount Siintis during the
years 1829-51,383 the other was carried out during the years 1863-78 by Prof.
A. von Kerner at the mountain ranges of the Inn Valley near Innsbruck. Obser-
vations at the Inn valley were done separately for the northern and the southern
slope. 384 We have compiled the significant results below, omitting the data for
January and February. In these months the snow line sank far below the mea-
suring station to some point in the valley and often could no longer be deter-
mined. The month of August was also excluded because during this time the
hillsides and peaks often have no snow at all and consequently no snow line
could be established.
383 Results from these observations are published in this Journal, Volume XVII, p. 49.
384 Compare the publication of F. von Kerner: Untersuchungen tiber die Schneegrenze im
Gebiet des mittleren Innthales [Observation of the Snow Line in the Area of the Central
Inn Valley]. Denkschriften der mathematische naturwissenschaftliche Classe der Wiener
Akademie der Wissenschaften. Volume LN Wi en 1887.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 195
Altitude of the Bottom Snow Line over Several Months
Inn Valle~ Mount Siintis
South Slope North Slope
m m m
March 960 720 720
April 1270 1110 910
May 1700 1540 1310
June 2190 2030 1910
July 2680 2470 2530
August
September 3210 2760 2470
October 2150 1890 1740
November 1300 1010 1020
December 740 680 740
If similar observations were available for a number of other locations in the
Alps, especially in the West and South, we might have the answer to the inter-
esting question how large the snow-covered areas are each month. We would
then have an essential database for the question about the influence of the snow
cover on the climate of the Alps.
Of all the official meteorological stations, only those of Bavaria regularly
observe the snow cover. The Royal Bavarian Meteorological Centre in
Munchen was the first among all the official meteorological centres in Europe
to be credited for having implemented mandatory snow cover observations at
their stations.385 This observation is very extensive indeed. The stations do not
merely note down daily if there is snow on the ground or not, but take an actual
measurement of the depth of the snow cover by using gauges that have been set
up at suitable locations. The observations began in the winter of 1886/87. If
they are continued for some years to come, it will be possible to determine
exactly the average altitude of the snow line of the Bavarian Alps, at least for
those months when the snow line extends below the height of Mount
Wende/stein, i.e., from October until the end of May. In looking at the
observations of the stations at Miesbach (717 m), Hochkreuth (1989 m) and
Wende/stein (1730 m) done during the five winters of 1886/87-1890/91, I
discovered that in the vicinity of Mount Wende/stein the bottom snow line
averaged at above 1700 m in October, at around 1000 m in November, below
900 m from December to March, at 1000 m in April, and at 1700 m in May.
Despite this very brief observation time the results tie in closely with those
gained at Mount Siintis and in the Inn Valley.
Even though we currently know little about the extent of the snow cover
during the various months, we are able to draw a number of conclusions re-
garding its influence on the climatic conditions of the Alps, especially in the
385 Only the observations of the Society of Science of Katharinenburg, Russia start at an
earlier date.
nico.stehr@zu.de
196 EDUARD BRUCKNER
winter. A conclusive inquiry, however, is out of the question because there are
not enough data available. Rather, the following remarks are meant to encour-
age further observations that could then later lead to a more detailed inquiry.
For the same reason, in describing the influence of snow I will focus on tem-
perature and humidity only. Before we, however, take this main step, we
should look at the physical attributes of snow that impose these conditions.
5.1 THE PHYSICAL CHARACTERISTICS OF SNOW
Air has the characteristic, especially when it is dry, to let warm rays from the
sun infiltrate without interference; the immediate warm-up during this process
is extremely low. The rays, therefore, reach the ground in nearly full strength
being eagerly absorbed by it and warming it up. One can witness this process at
any time in the mountains. If, on a mountain peak, air is measured away from
direct sunlight, its temperature will always be fairly low, whereas for example
the dark rocks and also the dark clothing of the mountain climber will be heated
up by the sun rays fairly quickly.
Warm air rises from the surface of the ground up into the air layers immedi-
ately above the surface. In 'many ways the surface functions like the top of an
oven. It is therefore of some significance with respect to the temperature of the
air whether the surface below heats up or cools down quickly or slowly.
Wherever the surface absorbs heat quickly, the air temperature will rise up
high; but where the surface heats up slowly and very little, e.g., as in the case of
a body of water, air temperatures will remain low. The same relates to the
cooling process: the lower the temperature of the surface, the greater the drop
of the air temperature.
This close correlation between the air temperature and the temperature of
the surface below remains, even when the ground is covered with snow. In this
case, due to the physical properties of snow, the temperature conditions of the
snow cover differ in many ways from temperature conditions of the bare
ground; this of course in turn influences the air temperature.
Snow never reaches a temperature above zero degrees. Once it has reached
a temperature of zero degrees and is warmed up further, this additional amount
of heat is used up entirely in the melting process. As long as the air temperature
remains above zero, snow will always be cooler than air and, consequently, will
be trying to cool the air by drawing on its warmer temperature for the melting
process. This is obvious and yet, as Woeikof rightly points out in his funda-
mental work about the influence of a snow cover on ground, climate, and
weathe~86 which we will have to quote more often, it is still being given too
386 Penck's Geograhische Abhandlungen, Volume III, Issue 2, p. 15, Wien 1889. Woeikof
was the first who drew attention to the significance of snow as a climatological factor.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 197
little consideration. This cooling effect of snow continues not only as long as
the thennometer indicates above zero degrees but also at temperatures below
zero.
To begin with, even at air temperatures below zero the snow cover is not
wanned by the sun as much as is bare rock or soil because a good portion of its
rays bounces off the snow surface without any wanning effect. This has to do
with the composition and colour of snow. If fresh-fallen snow is walked on,
crystals flash everywhere, as if the surface is covered with diamonds. All these
countless crystals reflect sunlight like a mirror, not just in one but in all
directions with the same kind of intensity. Reflection decreases as snow ages
because the crystals grow smaller during the melting process and snow
compacts even more. Even then the degree of reflection remains high. In
addition, the whiteness of snow, a result of its high volume of air, enhances
reflection. Zenker points out that soil reflects only 1130 of the infiltrating sun
rays while snow reflects 116.387 In my opinion this figure is much too low,
particularly for fresh-fallen snow. Just consider the strong effect of the heat and
brightness of sunlight reflected by snow, experienced when hiking across sunny
snow covered glaciers. Skin tans in no time and burns quickly. As white paper
already reflects approximately 40% of light, fresh-fallen snow should at the
very least reflect 113, that is at least ten times as much as soil. Regrettably,
exact tests are non-existent. All reflected sun rays are lost as they are now
increasing the temperature of the snow cover. It is for this reason that the
amount of heat available for this increase is much smaller in a snow cover than
in bare soil.
While, on the one hand, reflection causes the snow cover to wann up
more slowly, its enonnous capacity of radiating heat, on the other hand,
enables it to cool down rapidly.
How much heat a substance is capable of emitting depends, as is well
known, on the size of its surface. A surface reduced in size by a high polish
gives off the smallest amount of heat; in contrast, the rougher a surface is the
larger it becomes and the greater is the loss of heat. The surface of a snow
cover can be considered unusually rough even though it levels uneven terrain
and, therefore, appears smoother to the naked eye than bare soil. In relation to
its volume a single snowflake's surface is rather large with many edges which
enhance reflection. The snow cover consists of numerous such flakes; therefore
its surface space reflecting heat is many times larger than its actual size. This
enables the extraordinary cool-down during clear weather periods, particularly
at night. Radiative loss is highest after fresh-fallen snow allowing this type of
snow to cool down to the lowest temperature. The older the snow, the smaller
the surface becomes because of melting ice tips and compacting snow crystals
387 Vertheilung der Warme an der Erdoberflache [Distribution of temperature on the surface
of the earth], p. 63, Berlin, 1888.
nico.stehr@zu.de
198 EDUARD BRUCKNER
and subsequently its capability to cool down rapidly by giving off heat is
reduced.
To a large degree, the cooling effect on the snow surface is increased even
further because snow is a poor conductor of heat. As a result the amount of heat
lost in the top layer is not replaced adequately enough from below as is the case
with bare soil. The low heat conduction of snow is caused by its feather-like
structure capable of a large number of air pockets between the ice crystals.
Snow contains an inconceivable amount of air. It is easily calculated. One only
has to measure how much water a snow cover of a given density loses during
the melting process. Not until the past few years have such analyses been
carried out on a larger scale. However, there are almost none for the Alps. The
most complete series of observations was carried out by Abels at
Katharinenburg where snow-like in the Alps-often falls at fairly low
temperatures. 388 Abels discovered that the ratio of snow and melt-water from
newly fallen snow varies between 1:7.0 and 1:45.0; i.e., a new thick layer of
snow of 100 cm yields not more than 2.2 cm to 14.3 cm of water or a compact
layer of ice of only 2.4 to 15.9 cm thickness. P. Schreiber found similar data in
Chemnitz: 1:6.6 to 1:34.389 As a result, 83 to 97% of newly fallen snow is made
up of atmospheric air and only 3 to 17% of solid ice. The air content is always
higher, the lower the temperature and the weaker the wind was during the
snowfall. Snow is much more firmly packed in snow drifts where its air content
amounts to only 63 to 76%, according to Abels. Over time snow compacts,
then compresses. On the one hand it is probably the weight of the snow mass
itself that compresses it; on the other hand intervals of melting and freezing
obviously play an important role. Air is forced out during compression. Abels
discovered this while observing the density of the bottom layer of the snow at
the beginning of the winter, measuring it throughout the whole winter. The air
content dropped from 84% right after the first snow fall to 67% at the end of
the winter before the snow melted completely. The glacial mountain range
provides the opportunity of observing this process even further. According to
Ratzel, glacial snow of the Tyrol Alps has 55% air and that oflower layers with
less air cavities only 33%; the older the glacial layers, the lower the air content,
until it is reduced to a minimum fraction when glacial snow has turned into
glacial ice. 39O
Air is known to be a poor heat conductor. Consequently, the air content of
snow reduces the conduction of snow considerably, meaning that the snow's
conductivity improves the less air it contains. Because each layer of snow is apt
to lose air after it has fallen and begins to take on the consistency of glacial
snow and ice, conductivity increases with age.
388 Repertorium flir Meteorologie, Volume XV.
389 Meteorologische Zeitschrift, 1889, p. 141.
390 same journal, p. 433.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 199
The characteristic of snow being a poor heat conductor is well known by
every farmer. He has experienced many times that a blanket of snow prevents
cold temperatures from reaching the soil and protects the seeds from freezing:
The snow cover keeps the warmer temperature in the ground. This has been
verified by several measurements of the ground's temperature. The top layer of
snow may cool down considerably with no interference by warmer
temperatures from below which is evident from the fact that the temperature of
the snow increases rapidly from top to bottom layer. Based on observations
taken from both Becquerel, Woeikofarrived at an increase of 0.31 to 0.36 0 per
centimetre. Often the increase may even be much higher. On January 18, 1893
I measured the temperature of the snow on the terrace of the Swiss
Meteorological Centre in Zurich at 6 o'clock in the evening and got a reading
of -20.1 0 C at the surface, while 12 cm below the temperature was -6.10. This
results in a temperature difference of 14.00 per 12 cm, or 1.20 per centimetre.
The fact that the increase was not steady, but more rapid near the surface than
down below, leads us to the conclusion that as little as I cm below the surface
snow may be several degrees warmer than at the surface itself, proving how
effectively the surface layer prevents the cold from spreading. This is very
important for our investigation.
Let us summarise the properties of snow with regard to its influence on the
air temperature.
1. Snow never warms up to more than zero degrees and must therefore have
a cooling effect on air as soon as the air's temperature is above zero.
2. The surface of snow reflects a relatively large amount of the sun rays
and, consequently, cannot warm up as quickly as bare soil during frost.
3. The surface of snow is extremely well suited for reflecting heat and can,
therefore, cool off rapidly.
4. Because the heat conductivity of snow is very low due to its high air con-
tent, a loss in temperature cannot be made up for by warmer temperatures
from below.
5. The ability to reflect light and heat is strongest at the surface of newly
fallen snow and decreases with older snow. The opposite holds true with
regards to its ability to conduct heat. Consequently, under the same con-
ditions new-fallen snow will be colder than old snow.
5.2 OBSERVATIONS OF THE SNOW TEMPERATURE
ATDAVOS
As might be expected, temperature observations of a snow surface are of
great significance to the question of the influence of snow on the climate.
Unfortunately, up to this date they are only available for few locations. The
nico.stehr@zu.de
200 EDUARD BRUCKNER
most extensive series of measurements was carried out at Sagastyr, a polar
station at the mouth of the Lena River. Observations in the Alps have been,
until now, non-existent. At my suggestion, F. Imhof and C. Mosca from the
Meteorological Station at Davos kindly agreed to undertake measurements of
this kind for the months February, March, and December 1891 and January and
February 1892. The results are highly interesting.
The observations at Davos were carried out three times daily, at 7 o'clock in
the morning, 1 o'clock midday, and 9 o'clock in the evening on a softly sloping
meadow immediately behind the station,391 only a few meters from the building
and the shelter containing the station's air thermometer. This was located about
3 meters above the snow surface. The meadow had sunshine during the
afternoon, but was in the shade at the time of the observations. 392 That is why
the measured temperatures reflect only in part the snow conditions of an open
area of snow: The I o'clock-readings on clear days are too low in comparison to
a snow area exposed to the sun. Thus, in the course of our investigation we
have to keep in mind that readings relate to an area of snow in the shade.
Morning and evening readings are not affected by this; because as I learned
from my experiments on the terrace of the Meteorological Centre in ZUrich and
in my own garden in Bern the temperature of an area of snow is the same
shortly after sunset, regardless whether it has been exposed to shade or bright
sun light during the day.
First, I will refer to the monthly averages of the air temperature, the snow
surface as well as the difference between the two. The plus-sign indicates that
the air temperature was higher than the temperature of the snow surface, the
minus-sign indicates the opposite. The average cloud cover is also indicated in
tenths ofthe total area of visible sky.
Throughout the entire observation time the temperature of the snow surface
was lower than the air temperature on an average of 3.9°, and in none of the
observed months did the air temperature differ by less than 2.2°. The difference
was always greater at 9 p.m. than at 7 a.m. and the greatest at 1 p.m. The latter
391 Unfortunately thislocation was unavoidable for external reasons.
392 It must be noted that the obtained temperatures of the snow surface represent maximum
values, i.e., they are slightly higher than the actual temperatures. This is already indicated
by the fact that the thermometer repeatedly shows above-zero-degrees for the temperature
of the snow surface, which of course is not possible. Readings may be too high for the
following reasons: Firstly, it is not entirely unavoidable that the thermometer sinks into the
snow; consequently the instrument does not measure the temperature of the surface but of
the top layer, which may be several millimeters in depth. Secondly, the thermometer
protects the surrounding snow a bit from losing temperature. In any case, the loss in
temperature that the glossy ball of the thermometer undergoes is smaller than the
temperature gain of the surrounding snow particles protected by the instrument. Melloni,
on the other hand, discovered that glass is nearly as sensitive to temperature as soot. But
new experiments by H.A. Hazen, W. Koppen et al. have shown that a thermometer's
sensitivity is much lower than that of glass; it is between that of glass and that of metal.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 201
is explained by the fact that the snow area remained in the shade well past
1 p.m. If we omit the 1 o'clock measurement and use only the averages of the
7 a.m. and 9 p.m. measurements,393 we arrive at an average difference of 3.2 0
between air and snow temperature.
Monthly Averages of Air Temperature, Snow Temperature, the Difference Between the Two,
and ofthe Cloud Cover at Davos
7ha.m. I hE.m.
TemEerature Cloud TemEerature Cloud
Da~s Air Snow Diff. Cover Air Snow Diff. Cover
1891 Feb. 28 -12.0 -16.7 4.7 2.2 1.9 -7.2 9.1 1.8
March 31 -5.3 -7.0 1.7 6.6 2.9 -l.l 4.0 6.4
Dec. 394 12 -9.3 -13.9 4.6 4.5 -3.1 -10.1 7.0 4.0
1892 January 31 -9.7 -11.9 2.2 6.2 -\.7 -6.0 4.3 5.6
Feb. 395 20 -6.0 -7.5 1.5 6.6 1.2 -3.2 4.4 6.6
Average 122 -8.5 -11.4 2.9 5.2 0.2 -5.5 -5.7 4.9
9hE·m. Average 1/4 {7+l+2x92
TemEerature Cloud TemEerature Cloud
Da~s Air Snow Diff. Cover Air Snow Diff. Cover
1891 Feb. 28 -9.1 -14.7 5.6 l.l -7.1 -13.3 6.2 1.6
March 31 -4.0 -6.0 2.0 5.5 -2.6 -5.0 2.4 6.0
Dec. 394 12 -8.5 -14.0 5.5 2.5 -7.4 -13.0 5.6 3.4
1892 January 31 -8.8 -11.6 2.8 4.3 -7.3 -10.3 3.0 6.6
Feb. 395 20 -6.2 -7.7 1.5 4.2 -4.3 -6.5 2.2 5.4
Average 122 -7.3 -10.8 3.5 3.5 -5.7 -9.6 3.9 4.6
This is a greater difference than has ever been found previously. At
Sagastyr in the Lena estuary and at Katharinenburg, the only stations where
snow temperature 3% measurements were taken over a number of months, the
average difference between snow and air temperature from the 7 a.m. and
9 p.m. measurements was only 1.8 0 C, i.e., not much more than half of what
we found in Davos. Evidently, in alpine valleys such as Davos, the dif-
ference between air temperature and snow temperature is much greater than
in lower regions. The reason is quite obvious. Temperature is transmitted
much stronger at higher altitudes than in the plains because the protective
layer of the atmosphere is much thinner and contains less moisture. A second
reason will be introduced later.
393 This figure is of course not the average of the entire day.
394 The first 19 days of December are missing.
395 9 days are missing when thermometer readings were not taken because of a severe snow
storm.
3% The measurements of Christoni at Modena are not applicable because they are not taken
from the very top of the snow surface but 2 cm below the surface.
nico.stehr@zu.de
202 EDUARD BROCKNER
All in all, only 47 individual measurements, i.e., 13% of the total, showed
an air temperature equal to or lower than that of the snow surface. All of
these 47 measurements were taken on days with snowfall and, as a matter of
fact, according to the daily entries in the log book, 43 of them were done
while it was snowing. Consequently, at Davos the snow surface is warmer
than air only when it snows.
This result prompts us to take a closer look at the difference between air and
snow temperature of those measurements taken during a snowfall. This is
shown in the following table.
Temperature of Snow Surface and Air during Snowfall
7h a.m. (30 Observations) Ih p.m. (21 Observations) 9h p.m. (24 Observations)
Temperature Tt!lllperature ___ ~elllQerature
Snow Snow Snow
Air Surface Diff. Air Surface Diff. Air Surface Diff.
-6.3 -6.2 -0.1 -2.2 -\.7 -0.5 -5.6 -5.3 -0.3
Even the averages show the temperature of the snow surface to be warmer
than air during a snowfall. The result remains the same if a larger sample is
taken into account. Of all the measurements during a snowfall the snow surface
was warmer than air at 7 a.m. in 53% ofthe cases, at 1 p.m. in 62%, at 9 p.m. in
60%, and on average in 58%. This leads to the interesting result that at Davos
the temperature of falling snow is usually higher than the temperature of the air
near the ground it falls upon. As falling snow obviously has the same tempera-
ture as the air layers it originates from, the air temperature at the bottom of the
Davos valley is colder than air layers high above the valley. This can only be
explained by the fact that at the onset of a snowfall cold air stagnates in the
Davos valley whilst the warmer west winds above carry the snow. This conclu-
sion awaits of course confirmation by data from other stations.
If we compare the average temperature difference between snow and air
for each month as shown above with the average cloud cover, we clearly
recognise the strong correlation between this factor and the temperature of
the snow surface. We anticipated this because the cloud cover is mainly re-
sponsible for the intensity of the sun's influence. This influence is even more
evident if we classify the measurements according to the proportion of
overcast skies and determine the average temperature differences at those
various degrees of cloud cover. This is shown in the following small table.
The measurements taken during snowfall, already been summarised above,
have been omitted in this table.
7ha.m.
Temperature
Cloud cover No.ofobserv. Air Snow surface Difference.
o 28 -14.3 -20.4 6.1
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 203
7ha.m.
TemEerature
Cloud cover No. of observ. Air Snow surface Difference.
1-4 20 -4.5 -5.9 1.4
5-S 22 -4.3 -6.S 2.5
9-10 20 -4.5 -5.9 1.4
Ihp.m.
TemEerature
Cloud cover No. of observ. Air Snow surface Difference.
0 31 0.0 -10.0 10.0
1-4 19 2.2 -1.0 3.2
5-S 26 I.S -1.S 3.6
9-10 19 2.2 -1.0 3.2
9hp.m.
TemEerature
Cloud cover No. of observ. Air Snow surface Difference.
0 49 -9.7 -15.5 5.S
1-4 IS -3.6 -4.9 1.3
5-S 13 -4.3 -6.7 2.4
9-10 IS -3.6 -4.9 1.3
This table emphasises two facts. First, it shows the unusually close relation-
ship between the temperature of the snow surface and the cloud cover. On
bright, cloudless days when nothing interferes with radiation, the snow surface
is at its coldest; the more overcast the sky becomes, the higher the temperature
rises. The temperature of the snow surface takes on the temperature of the air,
but not entirely; in fact the temperature difference between the two increases
the brighter the sky is. As a result on overcast days the snow surface on an
average is only 1-3° colder than the air but on clear days 6-10°. This difference
increases at some morning and evening measurements to 12°, around noon
even to nearly 14°. I must admit that I had not expected such big differences.
They are by far larger than what has ever been measured in the plains.
To explain such enormous temperature differences in bright weather we
need to consider the overall atmospheric conditions. If skies are clear above the
Alps, that region is generally under the influence of a high-pressure system.
This was also the case during the time of our observations. Of the 108 days of
observations under a clear sky, during 95 days the Alps were exposed to
maximum barometric pressure, or anti-cyclonic conditions, to use another
expression. Under such weather conditions we have the strange phenomenon
which remained puzzling for a long time and has only recently been explained
by Hann, that the temperature increases from the floor of the valley towards
considerable altitudes. In winter, high pressure carries warm air to high
altitudes and cold air to the valleys. This is linked to the general circulation of
air in an anti-cyclonic system and to resulting thermo-dynamic conditions of
nico.stehr@zu.de
204 EDUARD BRUCKNER
the ground surface. In an anti-cyclonic system air moves downward: air masses
fall from high altitudes down toward the ground where they spread horizontally
in all directions. Although they come down from an altitude of many
kilometres up where their temperature was very low, they arrive near the
ground with their temperature considerably higher because while descending
the air is exposed to increasing amounts of pressure and is compressed; of
course, compression generates heat. This warming process is important: it
amounts to 10 C for a drop of 100 m. One is able to feel this very distinctly on
the peaks of the Alps where the air tends to be extraordinarily warm under anti-
cyclonic conditions, as Hann has pointed out previously. The same is not felt at
all at the floor of the valleys; on the contrary, here an icy cold prevails. This
cold is a consequence of the strong radiation from the surface of the ground
which is enhanced considerably by a snow cover. This effect is enhanced by
dry air and a cloudless sky, both characteristics of anti-cyclonic conditions. The
lower the humidity in the air, the less likely is any interference with radiation.
The cold temperature of the underground is passed on to the bottom layers of
air; for these do not circulate with the rest of the air, particularly when located
in secluded valleys-they stagnate. The more stagnant this air becomes the
stronger is evidently its ability to adopt the low temperature of the
underground-in our case -the snow cover-and the smaller is the difference
between the two temperatures. This is exactly what happens in Davos.
Whenever it is severely cold in the valley, it is considerably warmer in higher
altitudes. This is a known fact to locals and visitors despite the lack of
meteorological observations for these altitudes.
It is evident from our fmdings that in bright weather, i.e., under maximum
barometric pressure, the air at a certain altitude must be much warmer than the
snow surface below. What seems remarkable, however, is that in Davos this is
the case at as little as 2-3 m above the snow surface. If air would stagnate
completely in this location, the snow surface would have to transmit much
more of its low temperature to the bottom layers of the air mass so that there
would be only a small difference between air and snow temperature at a height
of 2-3 m. As we know from our fmdings that is not the case. We must
therefore conclude that the air in the Davos valley travels, if only slowly, and is
thereby replaced by descending warmer air masses. This slow air movement
can be well and quite reasonably explained because of Davos' location right
next to a very flat water shed with the valley of the Landwasser River running
to the south-west and that of the Landquart River (or a small side arm of this
river) to the north-west. And indeed, log book entries confirm that, on days
when Davos is under the influence of a maximum high pressure system, a very
light stream of air is noticeable travelling steadily at near-zero speed in the
direction of the Davos valley. We are able to state: The big difference in air and
snow temperature in Davos on bright days and anti-cyclonic weather conditions
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 205
is the result of an additional supply of wann air descending from higher
altitudes and replacing the slow stream of air moving along the valley. That is,
under anti-cyclonic weather conditions Davos is comparable to a slope
location.
The observed big difference between snow and air temperature at Davos
during anti-cyclone weather conditions refutes the statement made by several
sources that the cold temperature in deep valleys is, at least to a certain extent,
caused by air masses which cool off at night along the slopes and descend to
the floor of the valley. If in fact the snow surface along the slopes-and not the
surface in the valley-was to be the primary cause of the cold temperature of
the air in the valley, this air obviously should not be considerably warmer than
the snow surface at the floor of this valley. Consequently, the observations in
Davos show that the cause of the cold air at the valley floor is to be found in the
low temperature of the snow surface in that location and not in the snow
surface along the slopes. I have to admit, it certainly does not quite make sense
to me that air along the slopes which is known to be warmer during the day as
well as at night than air at the valley floor should carry cold air down into the
valley. All it can possibly carry is a wanner temperature because it is warmer to
begin with and wanns up even further by approximately 10 C per 100 m when
descending. From the observations at Davos we can conclude that under anti-
cyclonic weather conditions the cause for those extreme cold temperatures at
the floor of alpine valleys is a much more localised phenomenon than had been
previously assumed.
From the above discussion we may conclude further that under calm anti-
cyclonic conditions the temperature difference of air and snow along most
slopes will have to be especially pronounced because a steady stream of new
warm air flows down from higher altitudes replacing the old air falling to the
valley floor. In this case, air has even less time to cool down while moving
along the cold snow surface than it does in Davos. Observations supporting this
assumption will yet have to be carried out, of course.
It appears as if our explanations would contradict Woeikofs fmding that a
strong air turbulence reduces the temperature difference between air and snow
by increasing the snow temperature. But this only seems to be so: for, slow
hardly noticeable movements of air must have a different effect than strong
winds. This is simply the result of the lower specific temperature of air. We can
distinguish between three different scenarios in which an air mass comes into
contact with a snow surface of different temperature. If air is completely stag-
nant, a gradual transfer of temperature between air and snow must result in a
complete adjustment. In fact the air will cool down to the snow temperature
because its specific temperature is low. By comparison the difference in tem-
perature will be very small. If air circulates violently above the snow surface, if
for instance there is a strong wind, large air masses are rapidly being moved
nico.stehr@zu.de
206 EDUARD BRUCKNER
across the snow surface so that the snow almost reaches the air's temperature
despite the latter lower specific temperature. Again, in this case the temperature
difference is minimal. If however air moves very slowly across the snow, there
will not be sufficient time to adjust completely to the snow's temperature; but
at the same time, it is incapable of raising the snow's temperature substantially
because due to the slow speed the respective air mass being moved across is
small. In this case the temperature difference will be big.
Unfortunately, the Davos observations can confirm this thesis only in part
because complete air stagnation does not seem to happen in Davos where slow
circulation of air appears to prevail all the time. Neither are strong winds likely
to occur due to its sheltered location. Observations under a clear sky in the
morning generally show a zero wind force. Among the observations in the
evening (at 9 p.m.), however, four were made at a wind force of 0 to 1 and at 1.
As small as this number is, I still think it can be averaged and compared
with the overall average.
Observations Teml2erature °C
Number Air Snow Surface Difference
All Observations 49 -9.7 -15.5 5.8
Wind Force 0-1 Observ. 4 -5.6 -9.2 3.6
Difference (in 0c) 4.1 6.3 2.2
The temperature difference between snow and air was indeed small in these
four observations, i.e., on the average 3.6°, while all 9 p.m. observations under
clear skies show an average difference of 5.8°, which is 2.2° higher. This smal-
ler difference is primarily the consequence of the high temperature of the snow
surface, which is 6.3° higher than the overall average. Even the air is warmer
but only by 4.1 0. But this is exactly what is to be expected if we assume that a
stronger air turbulence warms up the snow surface. Strong movements of air
increase the temperature of the snow surface and reduce the difference to the air
temperature. The latter part of this statement corresponds with Woeikofs ob-
servations at the Lena estuary.
The observations of the snow surface temperature at Davos allow us to
approach another question that has, been raised recently, i.e., the possible
influence of the snow surface on the humidity of air.
It is generally assumed that, in the mountains, evaporation of snow is high.
That snow layers lose depth as they age is generally seen as proof of this view,
however this is not altogether correct. This loss is at least in part nothing more
than the result of compression when air is forced out during melting and re-
freezing. There is no evidence at all that high evaporation is the only factor in
reducing a snow cover and I am quite sure that such a generalised assumption is
incorrect. Much rather, the reaction of snow to the gas molecules of moisture in
the air is not always the same but alternates indeed. At times moisture is
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 207
extracted from the air and deposited as ice on the snow surface drying the air,
so to speak. At other times water is lost through evaporation raising the
moisture content of the air. The question is which of the two processes
outweighs the other. Woeikofbelieves that evaporation plays a greater role than
condensation and P.A. Muller had tried to prove it for Katharinenburg. In the
Alps, at least in the valleys, the situation is different as we have learned from
the observations at Davos.
This question can be solved by determining the dew-point of the air and
comparing it with the temperature at the snow surface.
If an air mass that contains a certain amount of moisture cools down gra-
dually, water will suddenly condense in liquid or solid form when a certain
temperature is reached-condensation occurs. This temperature is called the
dew-point. The more moisture is in the air, the higher will be the dew-point,
and the lower the moisture content the lower the dew-point. If we expose a
substance with a temperature lower than the dew-point of the air mass under
examination to this air, moisture will condense on the surface of the colder
substance, i.e., the substance will be covered with small dew-like water
drops or small ice crystals. If the temperature of the substance is known and
also the dew-point of the air, which is easily determined from the observed
moisture, we can predict whether condensation at the surface of that
substance will or will not occur. The question whether a snow surface at a
given moment condenses moisture taken from the air or gives off
evaporation, can, therefore, be put this way: Is the snow surface temperature
below or above the dew-point of air? If it is below, condensation takes place,
if it is above, evaporation.
For each single measurement of the snow temperature I have, based on ob-
servations of the moisture level in the air at Davos, determined the equivalent
dew-point in two different ways: Firstly, according to the old tables of dew-
points as they have been in use exclusively so far, and secondly, on the basis of
those figures which Friedrich and Ekholm point out as indicators of the dew-
point of ice.397 Limited text space does not allow for a reprint of these long
tables in their entirety; I shall contend myself with the averages for the
individual months and the observation times in the table below. It contains the
average snow temperature as well as the difference between dew point and
snow temperature. A positive difference means condensation at the snow
surface, a negative-evaporation.
397 The dew point of ice is slightly higher than the dew point of water.
nico.stehr@zu.de
208 EDUARD BRUCKNER
Relation between the dew-point of air and the snow temperature
7ha.m.
Number Snow Difference Against the Dew Point
ofDa~s Temperature Above Ice Above Water
1891 Feb. 28 -16.7 3.8 3.2
March 31 -7.0 0.1 -0.01
Dec. 12 -13.9 4.0 3.6
1892 Jan. 31 11.9 1.3 0.9
Feb. 20 7.5 -0.7 -l.l
AVll· 12 11.4 1.3 1.0
Ihp.m.
Number Snow Difference against the Dew Point
ofDa~s Temperature Above Ice Above Water
1891 Feb. 28 -7.2 2.0 1.8
March 31 -l.l -2.6 -2.8
Dec. 12 -10.1 32 2.9
1892 Jan. 31 -6.0 0.6 0.4
Feb. 20 -3.2 -1.8 -2.0
Avg. 12 -5.5 0.1 0.0
9hp.m.
Number Snow Difference allainst the Dew Point
OfDa~s Temperature Above Ice Above Water
1891 Feb. 28 -14.7 4.5 4.1
March 31 -6.0 0.0 -0.2
Dec. 12 -14.0 4.8 4.4
1892 Jan. 31 -11.6 1.9 1.4
Feb. 20 -7.7 -0.9 -1.2
Avg. 12 -7.3 1.9 1.5
The table shows that the dew point of air, whether it was determined in one
way or the other, is on the average higher than the temperature of the snow
surface, meaning that condensation occurs more often at the surface than
evaporation. This is particularly true in the morning and in the evening. Nega-
tive differences, i.e., evaporation predominates, are found in only two months,
March 1891 and February 1892. Both months had heavy snowfalls and
generally cloudy skies. In February and December 1891 in contrast, when snow
temperatures were below the dew-point by an especially high margin, weather
was consistently bright as happens quite often in the Alps during the winter
when they are under the influence of a high-pressure system.
These results are further confirmed by the following small tabulation indica-
ting how frequently evaporation occurred during the different months and times
of the day. I restricted my calculations to the dew-point under freezing condi-
tions.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 209
Freguenc;r of Snow EvaEoration
Days 7 a.m. 1 E.m. 9E·m. Average
1891 Feb 28 11% 11% 7% 10%
March 31 61 81 55 66
Dec. 12 25 33 8 19
1892 Jan. 31 32 57 27 39
Feb. 20 60 79 59 65
Average 122 39 53 31 41
Difference between the Dew Point of Air and the Snow Temperature at Various Degrees of
Cloud Cover
7ha.m.
Overcast Cloud No. of Snow
Formations Observations Temp. Difference against the Dew Point
Above Ice Above Water
0 28 -20.4 5.3 4.5
1-4 22 -15.7 3.1 2.5
5-8 22 -6.8 0.5 0.0
9-10 20 -5.9 -0.2 -0.4
Snowfall 30 -6.3 -1.1 -1.4
IhE·m.
Overcast Cloud No. of Snow
Formations Observations Temp. Difference against the Dew Point
Above Ice Above Water
0 31 -10.0 4.1 3.9
1-4 23 -8.1 2.6 2.5
5-8 26 -1.8 -3.3 -3.5
9-10 19 -1.0 -2.3 -2.5
Snowfall 21 -1.7 -3.0 -3.1
9hE·m.
Overcast Cloud No. of Snow
Formations Observations Temp. Difference against the Dew Point
Above Ice Above Water
0 49 -15.5 4.9 4.3
1-4 25 9.6 1.1 0.9
5-8 13 6.7 0.2 0.0
9-10 8 -4.9 -0.5 -0.7
Snowfall 24 -5.3 -1.3 -1.5
Freguenc;r of EvaEoration of the Snow Surface Determined on the Basis of the Dew-Eoint
Overcast 7ha.m. IhE·m. 9h. a.m.
above ice above water above ice above water above ice above water
0 4% 4% 6% 6% 0% 2%
1-4 5 5 16 22 28 28
5-8 45 55 89 92 46 46
9-10 55 65 84 84 62 62
Snowfall 80 87 90 90 75 79
nico.stehr@zu.de
210 EDUARD BRUCKNER
In February 1891, from a total of 100 observations, only 10 showed evapo-
ration and a total of 90 condensation, in December only 19 and 81 respectively.
In March 1891 and in February 1892 in contrast evaporation was noticed du-
ring 2/3 of all observations. On the average, in the morning and in the evening
the number of cases with condensation was twice as high as those with evapo-
ration, while at noon this number was almost equal.
Of all observations combined, an average of about 41 % indicates evapora-
tion and 59% condensation. If the dew-point of water is used instead of ice,
these figures vary a little. The number of observations relating to evaporation
increases to 44% and the one relating to condensation decreases to 56%. Again,
according to these numbers, condensation prevails.
As we have seen from the monthly averages, the cloud cover is a significant
factor in how the snow surface relates to moisture in the air. The following data
make this even more evident showing the difference between dew-point and
snow temperature on the one hand and the percentages of those instances with
evaporation on the other hand in relation to different degrees of cloud cover.
As long as less than half of the sky is overcast, condensation outweighs
evaporation c<;>llsiderably and even more so the brighter the weather is. At a
cloud cover between 5 and 8, the two processes are counterbalanced. At
completely overcast skies (clouds 9 to 10) evaporation outweighs condensation
quite considerably, but even more so during snowfall.
We are at the end of our discussion of the observations relating to Davos.
Let us once again summarise briefly the most significant results:
1. The temperature of the snow surface at Davos was consistently lower
than the temperature of the air. Only when it snowed was it the same or
higher.
2. The smaller the cloud cover, the lower is the temperature of the snow
surface. In addition, the difference in temperature between air and snow
increases as the cloud cover decreases.
3. As a consequence the snow surface has a strong cooling effect on the air
temperature which increases with a diminishing cloud cover. Thus this in-
fluence is most intense under anti-cyclonic weather conditions and lessens
with increasing wind levels when the snow surface adjusts to the tem-
perature of the swiftly passing air masses above.
4. The low temperature of snow often initiates condensation of moisture at
its surface in the form of frost crystals. This condensation is most frequent
in bright weather while under overcast skies evaporation is predominant.
The results from the Davos observations primarily apply to the alpine valley
of Davos. However, they are partly of such a universal nature that they un-
doubtedly apply to other alpine valleys as well-such as the Engadin. What
these relationships are like at the bottom of the mountain or at the peak on the
other hand cannot be predicted. Observations are the answer to that question.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 211
There can be no doubt, of course, that the cooling effect of the snow surface on
the air temperature will again be a factor in those locations.
5.3 INFLUENCE OF THE SNOW COVER ON
WEATHER CONDITIONS IN THE BAVARIAN
ALPS
From the Davos observations we learned about the quality of the snow cover's
influence on the temperature. Regrettably they can not tell us about its quantity.
This is only possible by comparing the temperature with and without a snow
cover present under otherwise equal conditions. It is not possible in Davos be-
cause there is snow on the ground each winter, and winter months without
snow are virtually non-existent. In order to make such a comparison, we have
to go to lower altitudes. The only location for a comparison of this kind are the
Bavarian Alps because observations of the snow cover are not being carried out
anywhere else.
Obviously a comparison can be made in two different ways. First, one could
compare two temperature readings of a particular location, one taken at a time
when snow was on the ground, the other when under similar weather condi-
tions and in the same season the ground was snow-free; or one could compare
the temperatures of two different locations simultaneously, where one location
has snow and the other doesn't. We will look at both methods.
In December 1888 and in January 1889, the region of the Alps, Bavaria and
Switzerland in particular, was under the influence of anti-cyclonic weather
conditions, i.e., the air pressure was high and strong air turbulence was absent.
The valleys were quite often fogged in, while higher elevations had dear skies.
The air pressure decreased in February and weather conditions changed
bringing plenty of precipitation. In December 1890 the weather pattern was
quite similar. During most of the month the northern foot of the Alps was at the
centre or at the periphery of the anti-cyclonic system. This is also true for
January 1891 when weather conditions were only temporarily disrupted by a
short melting period between the 21st and 25th. Even more pronounced were
the anti-cyclonic conditions in February 1891. However, in one point both
winter seasons differed considerably-in the amount of snow on the ground. In
the winter of 1888/89 until the end of January only a few spots in the valleys
were covered with snow. A permanent snow cover did not materialise until
February. In the winter of 1890/91, on the contrary, the ground was completely
covered with snow from the end of November onward. This development is
best reflected in the data for the beginning and the end of a permanent snow
cover in Bavaria as shown below.
nico.stehr@zu.de
212 EDUARD BRUCKNER
Duration of Permanent Snow Cover of the Bavarian Alps
Height Above Winter 1888/89
Station Sea Level (m) Beginning End Days
Miinchen 526 21.II IO.III 18
Rosenheim 446 3.II 24. III 50
Miesbach 717 3.II 12.IV 69
Oberstdorf 842 3.II 20.IV 77
Peissenberg 994 3.II 21.IV 78
Hochkreuth 989 3.11 16.IV 73
Wendel stein 1730 3.II 22.IV 79
1890/91
Station Sea Level (m) Beginning End Days
Miinchen 526 25.xI 22.II 90
Rosenheim 446 25.xI 5.III 101
Miesbach 717 25.XI IO.III 106
Oberstdorf 842 24.xI 22.IV 150
Peissenberg 994 25.xI 17.I11 113
Hochkreuth 989 25.xI 9.III 150
Wendel stein 1730 17.X 8.V 204
In 1888/89 there had been an occasional snow cover prior to the recorded
months but snow was not very deep and disappeared quickly as the following
figures demonstrate.
Number ofDa~s with Snow and the De~th ofthe Snow ~cm~
December January February
Da~s DeEth Da~s DeEth Da~s DeEth
88/89 0 0 13 6 24 13
Miinchen 526
90/91 31 II 31 36 22 21
88/89 1 1 17 10 26 25
Rosenheim 446
90/91 31 7 31 40 28 41
88/89 12 2 16 9 26 62
Miesbach 717
90/91 31 15 31 57 28 67
88/89 0 0 12 II 27 85
Oberstdorf 842
90/91 31 17 31 73 28 67
88/89 6 2 14 II 26 92
Peissenberg 994
90/91 31 33 31 60 28 51
88/89 20 10 17 17 27 100
Hochkreuth 989
90/91 31 22 31 70 28 84
88/89 20 14 20 25 27 72
Wendel stein 1730
90/91 31 16 31 148 28 90
Differences in snow covers are quite extreme in December and January and
do not even out until February. What were the temperatures like during that
time? The following small table has the answer.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 213
Avera~e Monthl~ TemEerature in the Winter 1888/89 and 1891/92
December Janua!}: Februa!}:
1888/89 -1.8 -3.5 -2.8
Miinchen 90/91 -7.1 -6.2 -3.4
Difference -5.3 -2.7 -0.6
1888/89 -6.8 -4.6 -3.8
Rosenheim 90/91 -7.1 -7.6 -4.1
Difference -6.3 -3.0 -0.3
1888/89 -3.5 -5.0 -4.5
Miesbach 90/91 -9.1 -8.2 -5.3
Difference -5.6 -3.2 -0.8
1888/89 -1.4 -5.8 -5.3
Oberstdorf 90/91 -10.2 -9.8 -5.9
Difference -8.8 -4.0 -0.6
1888/89 -1.4 -4.6 -6.0
Peissenberg 90/91 -7.8 -7.1 -2.9
Difference -8.7 -2.5 +3.1
1888/89 -0.7 -5.4 -10.0
Wendel stein 90/91 -5.5 -9.9 -4.6
Difference -4.8 -4.5 +5.4
Clearly, the winter with heavy snowfall shows a lower average temperature
in December and January than the winter with less snow. This confirms our
finding that a snow cover reduces the temperature considerably. The months of
February 1889 and February 1891 cannot be compared directly because the
first one was stormy or "cyclonic", the second one "anti-cyclonic". The temper-
ature appears to be the same. However in February of 1889 it is the result of
high morning and high evening temperatures and in February of 1891 of high
temperatures at noon. If we exclude the latter and consider only the night tem-
peratures, i.e., the daily minima as shown below, February 1891 with a snow
base in the valley and in the plains was considerably colder than even January
1889 with hardly any snow, although normally February tends to be warmer
than January.
Average Temperature Minimum
January 1889 February 1891
Miinchen -5.5 -6.6
Rosenheim -6.7 -10.5
Miesbach -8.5 -10.2
Oberstdorf -1l.5 -11.7
nico.stehr@zu.de
214 EDUARD BRUCKNER
Conditions in Switzerland were similar; unfortunately no records about the
snow cover are being kept there. Only Zurich and Basel report the monthly
number of days with snow. In the winter of 1888/89 in the time period from
December 1 to February 1 Basel had a light blanket of snow on January 7 only,
and ZUrich from January 10 to 31. In Zurich however the snow cover was per-
manent in all of December of 1890 and January 1891, in Basel from Decem-
ber 1 to January 24.
TemEeratures in these months were as follows:
December Janua!1:
1888 1890 Difference 1889 1891 Difference
Basel -0.7 -4.5 -3.8 -1.4 -4.5 -3.1
Zurich -1.5 -5.8 -3.3 -2.5 -5.3 -2.2
The differences are more pronounced in December because during this
month in 1888 there was no snow at all, as opposed to the permanent snow
cover of 1890, while Zurich had only a partial snow cover in January 1889 and
on the other hand Basel was snow-free towards the end of January 1891. In any
case, these figures again demonstrate quite impressively the cooling effect of
snow.
But even when we compare the temperatures of two different locations, one
with snow and the other without, as they occur at a particular time of day we
clearly notice the cooling effect of snow. However, it is not easy to fmd two
such comparable locations, especially in the winter, a time when the region of
the Alps is normally under a permanent snow cover. Spring, however, offers
ample opportunities for comparison because wherever snow remains on the
ground, the temperature does not rise substantially above O. Since the moun-
tains get large amounts of snow, it often stays longer than at comparable alti-
tudes in the plains delaying the onset of spring. At a time when Bern has no
snow anymore and temperatures rise well above zero, the floor of the alpine
Aar valley at the same altitude is still buried deep in snow and the temperatures
are crisp.
This cooling effect of the snow cover is emphasised quite well by obser-
vations of the Bavarian stations. For comparison I chose two pairs of
stations, at Lindau and Kempten and at Rosenheim and Traunstein. In
Lindau at 399 m sea level, the ground is often free of snow, while in
Kempten located 300 m higher up (696 m) the snow still keeps. In the same
way Rosenheim (446 m) is often free of snow, while Traunstein (597 m) still
has snow. The temperatures of the paired stations cannot be compared
directly because the locations are at different sea levels and therefore, with
few exceptions Kempten will always be colder than Lindau, and Rosenheim
colder than Traunstein. If snow has a cooling effect, the temperature
difference between Lindau and Kempten and Rosenheim and Traunstein
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 215
respectively must obviously be particularly large as soon as the lower-level
stations are snow-free and the upper locations are not. However the
difference will be smaller if both locations either do or do not have snow at
the same time. This is indeed the case. Based on the snow observations of
the five winters from 1886/87 to 1890/91, I was able to identify those days
when snow conditions at both locations were equal, and those days when
only the upper locations had snow. The temperatures for those particular
days were provided by the data published annually by the Bavarian Central
Registry for those 4 stations and I came up with the following small table. 398
Temperature Difference Number Temperature Number Temperature
Between Kempten and Lindau of Dai:s Difference of Dai:s Difference
December Janua!:i:
When L. was snow-free,
25 3.6°C 25 3.6°C
K. had snow
Both locations had
100 3.1 110 3.2°C
or did not have snow
Difference 0.5 0.4
Februa!:i: March
When L. was snow-free,
30 4.l o C 40 2.2
K. had snow
Both locations had
80 3.0 45 2.0
or did not have snow
Difference 1.1 0.2
Temperature Difference
Between Rosenheim and Number Temperature Number Temperature
Traunstein of Dai:s Difference of Dai:s Difference
December Janua!1:
When R. was mainly free
IS 1.2°C 30 1.2°C
of snow, T. had snow
Both locations had
130 0.5 105 -0.4
or did not have snow
Difference 0.7 1.8
398 The analysis was made with pentade [5 days] averages of temperature. Pentades of equal
type were considered those pentades, when at both the upper and lower location, snow was
present or absent for at least 4 days. Pentades of opposite type were defined for Lindau and
Kempten as those when Lindau had no snow at all and Kempten all the time; for
Rosenheim and Traunstein, when opposing conditions prevailed for at least three days.
This was required, because the number of pentades of opposite type would have been too
few (only 8) for Rosenheim and Traunstein.
nico.stehr@zu.de
216 EDUARD BRUCKNER
February March
When R-was mainly free
20 1.5°C 30 1.7 °C
of snow, T. had snow
Both locations had
130 0.6 9.5 1.0
or did not have snow
Difference 0.9 0.7
During the 25 days in December when the snow cover in Kempten
remained, while it had already disappeared in Lindau, Kempten was colder
by O.5°C, during the 25 days of January by OAoC, and during the 30 days of
February even by 1. 1°C. These conditions were similar for Rosenheim and
Traunstein.
When in spring the snow cover has disappeared from the deep valleys and
starts retreating to high elevations, it still maintains its winter temperatures by
repeatedly cooling the air, especially at night, down to 0° and below. This
process intensifies the contrast between the snow-free valley down below
where spring has arrived already, and the wintry mountain valley. At that time
the difference in temperature between the two locations is at its highest.
Woeikofhas demonstrated this nicely by citing a number of examples from the
Swiss Alps.399 In April the bottom of the snow line is at an altitude of approxi-
mately 1000 m. The city of Chur (603 m) is already snow-free, while nearby
Churwalden (1213 m) or at least its neighbouring slopes still have snow which
does not disappear before May. Consequently, in April Churwalden is colder
than Chur by 4.6°C, in March by only 4.2°C and in May by only 4.3°C. In June
the snow line has moved up toward the 2000 m mnge and covers the area
between Sils (1810 m) and the Ju/ierpass (2244 m). As a result, the temperature
difference between both locations is highest in June: in May 2.5°C, June 3.7°C,
July 2.6°C. We can make the general statement that the temperature difference
between two locations diminishes most rapidly with increasing altitudes in the
month in which the lower location is snow-free while the upper location still
has some winter snow. This is especially true for higher elevated valleys, but
not for mountain peaks. These are towering in atmospheric conditions where
the contact between air and snow surface is very brief, where air is quickly
replaced and cannot cool down significantly. On mountain peaks as well as on
slopes, the temperature can, therefore, rise far above zero despite the presence
of snow; in valleys where air circulation is severely hampered, the average
temperature does not stray very far from zero degrees as long as there is snow.
This rapid temperature decrease in early summer is in some respect of im-
portance. Woeikof pointed out quite correctly that it may be one of the main
causes for the high number of heavy thunderstorms in early summer because
the rapidly falling temperature induces air to circulate upward carrying the War-
399 Woeikof, op. cit., p. 98.
nico.stehr@zu.de
ABOUT THE INFLUENCE OF SNOW COVER 217
mer moisture-laden air masses from the valleys. This causes heavy condensa-
tion and thunderstorms.
We are at the end of our discussion. In its course we had to be satisfied with
demonstrating the influence of the snow cover on the climate by giving at least
a few examples. Furthermore, these examples were taken from a fairly small
region, mainly between the Inn and the upper Salzach in the south, the Rhine in
the west and the lower Salzach in the east because the required observation
records were only available for this region. Although examples are of good use
whenever a problem needs to be illustrated to a larger audience-this, however,
is no way to find a defmite solution to the problem. At each step we were
confronted with new questions which could not be answered because the
necessary data are lacking. This means, observational data have to be collected
for all the regions of the Alps.
This is desirable in three directions. Firstly, it is required to determine the
annual vertical movement of the bottom line of a snow cover as has been done
at Mount San tis and near Innsbruck. Observations of this kind would be of
particular importance for the farthest East of the Alps, such as the lower Tauern
range, as well as for the southern valleys of the Alps. They can easily be carried
out by anyone, but have to be done daily or almost daily on a continuous basis
over a number of years in order to be meaningful. Hillsides that are not too
steep are best suited for this. The elevation of the snow line can easily be
estimated daily from a standpoint in the valley by using certain points on the
slope of which the elevation is known, for visual orientation. Prof. v. Kerner's
observations in the Innthal can serve as an example.
Secondly, it is necessary to make as many observations as possible at loca-
tions with meteorological stations using the same method of daily snow obser-
vations which proved so successful in Bavaria, i.e., recording day after day if
snow is on the ground or not, and possibly measuring its depth. This depth can
be determined simply by using measuring poles (Schneepegel) once daily in
snow as far away as possible from buildings, walls, trees, in short from all
obstacles which could cause snow piles. Again, such observations are very
simple and require very little effort and material. A manual explaining this
method has been published by the Central Meteorological Station in Bavaria.4oo
Thirdly and finally, it seems desirable that observations of the temperature
and the snow surface like the ones in Davos should be carried out in some other
locations, particularly in areas with a variety of different elevations. Of
particular interest would be observations in a location at the foot of the Alps, in
one of the steep main valleys as well as at a mountain observatory. These
observations require more effort and attention to detail as well as a certain skill
in handling sensitive thermometers. They should therefore best be undertaken
by meteorological stations.
400 Published in Meteorologische Zeitschrift, 1887, p. 15.
nico.stehr@zu.de
218 EDUARD BRUCKNER
Once observations have been carried out in all of these three directions and
published, one will be able to take the next step examining the influence of the
snow cover on the climate of the Alps in its entirety as we have outlined in our
remarks. They have served their purpose already if they entice one or the other
person to carry out observations of this kind.
nico.stehr@zu.de
Chapter 6
Influence of Climate Variability on Harvest and Grain
Prices in Europe *
Over the past 15-20 years the German farming industry went through a
severe crisis. A critical situation emerged that was unlike any other
before-caused by the stiff competition from American and particularly
Russian grain on the German market. This competition evolved within a few
decades, for not until the 1870s and 1880s did Russia's grain export increase
so enormously. Several events seem to be responsible for this boom in
Russian export: during the second half of the 1860s the North American War
of Independence paralysed America, Russia's biggest competitor. Expansion
of the German and especially the Russian railway system reduced transporta-
tion costs and instantly brought the rich grain fields of South and Southeast
Russia closer to the West. At the same time, Russia's agriculture expanded.
In addition, the exchange rate of the Rouble dropped offering a tremendous
incentive for export. Because of the low exchange rate while prices of goods
stayed the same abroad, the Russian trader had more money in his hands
than before; but above all the demand for grain in Central and Western
Europe increased with a rising industry and growing population. These are
the major reasons that have been held responsible for Russia's increasing
export. They can be attributed to man and his struggle for survival. Another
instance, however, which is completely beyond human control, should not be
omitted but rather be included in this number of causes despite its being
called secondary: It is the fact that Russia had a series of exceptionally rich
crops in the 1860s and 1880s up to and including the year 1888 and only
very few poor harvests, whereas the West suffered numerous crop failures.
This aspect has been neglected so far. One has become used to the fact
that poor harvests are considered chance occurrences whose economical
impact on a country may change the grain price and grain trade considerably
from year to year, but not regularly. In other words: those factors which have
Der EinjlufJ der Klimaschwankungen auf die Ernteertriige und Getreidepreise in Europa.
Geographische Zeitschrift, 1895.
219
nico.stehr@zu.de
220 EDUARD BRUCKNER
nothing to do with human intervention are considered to have no lasting
effect on the outcome of crops as soon as longer time spans, such as decades,
are concerned and not the individual year. There is, however, no justification
for this assumption.
I.
The outcome of crops depends to a large extend on the climate of the region
because it is the climate factor and not the soil which ultimately decides
whether a plant grows or not. The plant life on this earth develops therefore
in close dependence on the climate. Both, temperature and precipitation are
determining factors; the former is mainly responsible for defining the polar
boundaries of plants; the latter often determines their latitudinal expansion.
If weather conditions would remain the same year after year, the habitat of a
plant species would, ceteris paribus, be defined by the same isotherm and
the same isothete all over the globe. In reality this is not the case because for
plants to disseminate, more important than the average temperature and the
average precipitation is the variability of both elements. For several years a
tree may have had barely sufficient moisture to grow in a particular location,
it and its seedlings will perish if in the following year it is denied even this
bare minimum. This is a common fact. In a series of good years a plant will
perhaps extend its habitat; yet, the first bad year which does not supply the
warmth and moisture required for its survival will force it back into its old
boundaries. Plants are only able to survive in the long run in those places
where even in bad years the required minimum amount of warmth and
moisture is supplied. Obviously, weather conditions of bad years in
particular are responsible for limited plant dissemination. This primarily
refers to the natural habitats of plants.
Where man interferes, conditions are in part different. It can often be to
his advantage to grow a plant that prospers in good years only, but refuses to
bear fruit in bad years or perishes entirely. The yield of one good year may
compensate for many bad ones. For example, in medieval times and later
there used to be vine growers in many regions of northern Germany and
northern France, where viniculture is non-existent today, even though they
did not produce at all in some years. Because of the high freight costs it was
more advantageous to accept poor crops than to import grapes from the
south. As infrastructure improved later on and transportation became
cheaper and taste may have become more refined, the viniculture retreated
back to the south and vineyards were left to some other cultivation more
suitable for that climate.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 221
A similar development is seen happening today in the area of grain
production.
Temperature conditions determine the polar boundary of the various
types of grain cultivation. Temperature restricts wheat growing in Europe
primarily to regions south of the 60th or 61st parallel, yet permits the more
fastidious types of grain like barley and oats to advance almost up to the
Northern Cape and to the White Sea. Nonetheless in the main agricultural
regions of Europe the temperature's impact is limited to those regions and
therefore its effect on the outcome of the crop is only secondary. Only near
the grain's most northerly line of growth, e.g., in Scandinavia and Northern
Russia, crop is often damaged due to frost. Too much heat does not destroy
grain at all as long as enough water is available, e.g., wheat grows well in the
tropical part of India's subcontinent!
Moisture is much more important for growing wheat. Without water
there would be no agriculture, but with too much water none either. This fact
comes to mind when we look at the extent of agriculturally used land on this
earth or rather, examine the cause of crop failures. In all those regions where
rainfall is sparse, droughts and poor harvests go hand in hand. In regions
with excessive rainfall, bad harvests are mainly inflicted by rainy years. Of
course, it is not the absolute quantity of rainwater that tips the scale, but the
ratio between the amount of water and evaporation. The same amount of
water that inevitably drowns the crop in a cold climate may just be sufficient
enough for the irrigation of grain in a hot climate. In contrast, a small
amount of rain that could cause a drought in a hot country due to heavy
evaporation is often entirely sufficient in a cool climate. In Europe the wet
coastlines of the North Atlantic Ocean and Central Europe respond in a way
exactly opposite to the continent's dry interior. Southern Russia and Great
Britain including Ireland are extremes in this regard. Drought has to be the
cause of the crop failure in 1891 in the central and eastern provinces of
Russia. From August to October 1890, drought damaged the winter grain
crop; the following winter was dry with little snow so that the seedlings
froze in part. Due to the small amount of snow that melted early, the soil did
not get the usual amount of moisture. The frosts of April 1891 were not met
by a protective blanket of snow and damaged the seeds. Consequently winter
grain was destroyed. The summer grain however fell victim to drought and
hot winds in May, June, and July: crop failure was the result.401 Such events
are not too unusual in Russia, although they, fortunately, rarely reach such
enormous dimensions as in 1891.
In England, the exact opposite happened when numerous crops failed,
followed by high costs of living at the end of the thirties, during the forties,
and at the beginning of the fifties, as well as when harvests turned out poorly
401 Woeikof, in Meteoroiogische Zeitschrift, 1892, p. 40.
nico.stehr@zu.de
222 EDUARD BRUCKNER
in 1872, 1875, 1877, and 1879; all of which were excessively wet years.
Complaints were raised that because of extreme wetness in the fall, parts of
the agricultural lands could not be cultivated and in the wet summer grain
was flattened and rotted in the fields. 402
Central Europe and the German Empire in particular are in between these
two extremes. Dry years are usually good years for a number of agricultural
activities, especially for wine and fruit growing, but also for grain, whereas
they are bad for the grasslands and cattle raising. Although there are some
years when crops are damaged by dryness, however, more common are bad
harvests due to too much wetness. We are reminded of the poor harvests of
the wet years around 1880.
In Southern Europe and to a larger extent also in the tropics, at least as
far as the production of grain is concerned, conditions are similar to
Southern Russia. Years of famine in the East Indies coincide with dry years.
It would be of interest, at least for Europe, to determine the borders of
regions where grain growing is primarily damaged because of too much rain,
and those regions where damage is mainly caused by drought. A precise line
cannot be drawn however, because there is a transitional zone where
damages by too much rain are just as often as those by drought. In addition,
this borderline may well be different for each variety of grain. On the whole,
though, this line can be drawn from interior Russia to the Southwest toward
the Carpathian mountains and continuing along their range to the Alps all the
way to Southern France. In the East it still runs through the forest belt but
parallel to the steppes; in the West it almost coincides with the border of the
subtropical region. The area to the North suffers more from excessive rain,
the one to the South more from drought.
In those European countries influenced predominantly by ocean climate,
damages to grain cultivation by too much precipitation are so significant
that, given the present high wages and low freight costs inviting foreign
grain imports, grain production has to deal with the greatest difficulties.
Farmers have partially abandoned wheat growing in favour of grassland and
cattle raising. This can be said of Great Britain and Ireland, Holland,
Denmark, Scandinavia, and the provinces of Schleswig-Holstein and West
Prussia, also Switzerland, and, in general, the regions on the northern slopes
of the Alps which enjoy a lot of irrigation and not too much warmth. The
grasslands need water; it has to rain a lot before it becomes too much. Wet
years which would destroy the grain indigenous to the steppe are quite often
good years for raising cattle and vice versa. This has happened in the dry
summer of 1893: pastures, especially in the plains and foothills, in
402 Numerous examples are mentioned by Tooke and Newmarch, Geschichte und Bestimmung
der Preise 1723-1857 [History and Definition ofPrices 1723-1857], in German by Asher,
Vol. I and II.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 223
Switzerland and Southern Germany were dried out, whereas the wheat fields
had a good yield. The regions of Western Europe with cool summers are
indeed predestined for raising cattle and for grass cultivation. Here, this is a
natural line of production and the transition to cattle raising which is taking
place under pressure from competition by huge grain suppliers such as
America, Russia, and Hungary, is simply a more climate-conscientious way
of making use of the land.
If we move Southeast from our borderline the danger of poor growth
because of drought increases steadily. In particular, the regions east of the
lower and middle part of the Volga River suffer quite often from famines as
a result of drought. Here again, we have an outpost of grain growing regions
that are presently still viable because labour as well as land is cheap and
wheat cultivation therefore survives even multiple poor crops, but mainly
because a more natural and at the same time more advantageous use of the
land still has to be established, as has been done in the west combining cattle
raising and grassland cultivation.
My comments have been sketchy so far; a geography of the causes of
crop failures still has to be written. Presently, sufficient data are not yet
available. Data about harvest damages are not yet systematically collected
for all regions, as for example has been done in Prussia. Only some general
outlines could be mentioned here. Because of the variability of the climate
one weather system may be the most frequent cause of crop failures in one
location and another in some other location. Indeed, complete shifts in
agricultural production occur under the influence of climate, although they
are instigated by man-made constructions--by the system of modern
transportation: That the agricultural use of land has been given up in exposed
areas close to the ocean, is to a large extent the result of an adverse climate.
This inclemency has always existed, but in former times it was more
reasonable to cope with it, today it is more advantageous to accommodate.
Just as the climate can facilitate a local shift in agricultural production, it
can influence the timing of economical factors, especially the harvest and the
grain price. I will explain this briefly.403
403 A detailed explanation of this question must be left to a later larger publication because my
data collection is far from complete. Even here, I can only publish a portion of what I have
so far accumulated to avoid being too explicit. The following questions have been
addressed in several of my publications: Ed. Bruckner: Inwieweit ist das Klima konstant?
[How Constant is Today·s Climate?-Chapter 3 in this collection], Verhandlungen des
VIII. Deutschen Geographentages zu Berlin, 1889, p. 10 l. -Bruckner,
Klimaschwankungen seit 1700 [Climate Variability since 1700-excerpts are given in
Chapter 4 of this collection], Wien, Holzel, 1890, p. 275-279. -Bruckner, Ober die
praktische Bedeutung der Klimaschwankungen. [About the Practical Importance of
Climate Variability], Compte rendu du Vme Congres international des Sciences
geographiques. Bern 1892, p. 616. -Bruckner, Rufilands ZukunJt als Getreidelieferant
nico.stehr@zu.de
224 EDUARD BRUCKNER
II.
In view of the climate's strong influence on the outcome of the crop, one is
obviously only allowed to exclude the weather factor when explaining the
economical conditions and their change from one decade to the next, if it can
be proven that the average weather condition or, in short, the climate
remains constant. Only then the probability of good harvests in a region will
remain steady year after year.
Until recently, it was believed that the global climate is indeed constant.
New research has however shown that this is not the case; in fact climate
undergoes many changes which vary around a middle range. I do not refer to
the II-year weather cycle that has often been cited but not yet verified,
which from time to time disappears entirely and is caused by the II-year
cycle of sunspots, but rather to the far more important 35-year climate
variations. The highly regarded director of the Bavarian Meteorological
Institute, C. Lang, who, regrettably, passed away before his time, has
established their occurrence for the region of the Alps in 1885 in his usual
clear and precise way. I had the privilege of applying his findings globally.404
Meanwhile the obtained results have been confirmed by various sources,
particularly by R. Sieger, E. Richter, and Heintz whilst Fritsch, Sonklar,
Fore1 and Ellis had already confirmed their likelihood for certain specific
regions earlier. 405 Nowadays, climate variability has to be considered a fact.
Climate variations involve changes in temperature, air pressure and
rainfall that occur globally at the same time. The length of these variations,
i.e., the time that elapses between two extremes of the same kind, is 35 years
on an average, sometimes more sometimes less. The temperature is the one
factor in this process on which all the others depend upon. These
temperature variations are common to almost all the countries around the
globe. All of them experience cold periods as well as warm periods
simultaneously. The change in temperature averages 1°C. That is quite a lot,
meaning nothing else but that, for instance, during the five years around
1840, the average yearly temperature of Berlin was a full degree lower than
during the five years around 1825 which would be the same as if Berlin had
moved by 3 degrees latitude to the north.
Temperature variations have an impact on the distribution of air pressure.
The transfer of moist maritime air from the ocean to the continent seems to
[Russia's Future as a Grain Supplier], Supplement to Miinchener Allgemeine Zeitung,
November 19, 1894. Some sentences have been transferred from the latter two
publications with pennission from the respective editors.
404 Compare my book: Klimaschwankungen seit 1700 [Climate Variability since l70D-see
also Chapter 4 of this book], Wien, Holzel, 1890.
405 The complete literature about climate variability, summarized by myself, is to be found in
Geographisches lahrbuch, XV, 439 and XVII, 348.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 225
be hampered during warm periods, and easier during cold periods. This in
turn will influence the amount of rainfall in that region. Over most of the
land masses rainfall varies to such an extent that the cool periods are at the
same time moist and the warm ones dry. In our latitudes the amount of
rainfall during the wet period is 20% higher than during the dry period, in
Russia around 25-30% and in West Siberia even more than 100%.
Over the last two centuries cold periods with a lot of precipitation on the
continent seem to centre in the years 1705, 1740, 1775, 1815, 1850, and
1880; warm periods with dryness on the continent in the years 1720, 1760,
1790, 1830, and 1860.
Because of the wide range of these climate variations, they will have an
impact on the economy as well. That is actually the case; their influence on
the yields in agriculture-largely through rainfall-is significant. The
quantity and quality of vintages in France, Germany and Switzerland, by the
way, are better in dry and warm periods than in wet. It is more significant
that grain crops show the same impact.
Unfortunately, the available amount of data proving this impact is rather
limited and lacking in quality. Long-time observations of the yields of field-
produce in kilogram per hectare would be most suitable. However, data like
this are available only for the past few years. For Russia in particular, the
giant among grain producers, they do not go further back than 1883. The
situation is better in Prussia where the so-called Erdrusch-tab1e is available
since 1859 accumulating these data for each province and each year. If one
would have to rely exclusively on such harvest tables, there would not be
much point in comparing crop results and climate variability. Fortunately a
different indirect approach leads to results, at least for earlier decades-I am
talking about a closer look at grain prices.
Prior to a world trade in grain the price of grain was, first and foremost,
decided by the outcome of the crop in the producer country. Years with poor
crops have always been years of high food prices and vice versa. Therefore,
the annual changes in grain prices reflected changing crop yields. Though
the price was occasionally influenced by other events such as war or
pestilence, these factors were merely reducing the crop outcome's impact,
not obliterating it; the two kinds of impacts simply added up. We are,
therefore, entitled to draw conclusions from the development of the grain
prices of those earlier years to the crop yields. Lately, of course, this holds
no longer true. Nowadays grain prices are no longer solely affected by the
turn-out of a crop in the producer country but also by that of American and
Russian crops. Today, we have a world price for grain, which strongly
influences prices in individual countries. Therefore, only until the years 1850
or 1860, in England even up to 1840, can the grain prices in central and
nico.stehr@zu.de
226 EDUARD BRUCKNER
western Europe serve as an indicator for crop yields in the producer country.
With this restriction in mind, we can use these prices for our purpose.
The method of categorising the data is the same I have previously used to
determine the variability of meteorological factors. Five years each were
combined to a lustrum average, e.g., 1801-05, 1806-10, 1811-15, etc. This
way, random influences caused by irregular weather patterns from one year
to the next, were eliminated in most cases, yet long-term patterns remained
unchanged. Numbers obtained in this way are shown in the tables at the end
of this essay. They each contain, in addition, the amount of rainfall in
percentages of a multiple-year average. Where rainfall observations were
unavailable, I have included the dates of the grape harvest. I have previously
proven that these dates primarily follow the climate variations as they
change from lustrum to lustrum. The lustra averages of the rainfall as well as
of the dates for the harvesting of grapes are expressed in their difference to
the multi-year average; the minus (-) symbol indicates that the rainfall's
percentage was too low and the onset of the grape harvest too early by the
respective number of days. Therefore, the minus symbol generally indicates
a dry period. All relevant explanations as well as sources are included in the
tables' footnotes. The average yield and price for the wet and dry periods are
mentioned at the end of each table. These figures were arrived at by
determining firstly the time span of wet and dry periods for each country
based on the average rainfall per lustrum and then by determining the
average value for these time spans.
In addition, some of these numbers are represented graphically. Prior to
this, they were mathematically adjusted in order to exclude even more of
those random occurrences and, in doing so, to show their line of
development much better. I used the same method of adjustment as in my
research on climate variations according to the (a + 2b + c)/4 = ordinate of
b, where a, b, and c are neighbouring lustra averages. For the first
respectively last link, the formula (2a + b)/3 = ordinate of a respectively
(r + 2)/3s =_ordinate of s was applied.
The time spans marked by the average year of the lustrum at the top and
at the bottom, e.g., 1803 represents the lustrum 1801-05, were expressed as
abscissa and the figures for grain prices, yield, or rainfall as ordinates.
The rainfall curve rises and falls according to the amount of rainfall. An
up or down of the curve by one increment represents a change in rainfall by
2 112% of the multiple-year mean. The curves for the times of the grape
harvest, the prices, and the yields as well as Russia's rye exports again rise
and fall with the quantities they represent-but the scale is different.
Let us discuss these tables406 and curves.
406 Tables are placed at the end of this essay.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 227
The influence of climate variations on the harvest as well as on the price
of grain is obvious, but varies considerably in different regions. The same
difference between West and Central Europe on the one hand and East
Europe on the other, which we discovered earlier with regard to the cause of
crop failures, can be established again. During dry periods regions with
ocean-climate have high yields and consequently low grain prices because
their crop failures are mainly caused by too much rainfall, as in England,
France, Belgium, Denmark, Germany, and Austria. Regions with a continen-
tal climate, as for instance Ohio and Russia, have low yields and high prices
during those periods. In contrast, the yields of oceanic countries are low
during wet periods and prices high, whereas continental agricultural regions
enjoy good harvests and low prices.
Changes in yields are quite pronounced in Prussia. Whenever the amount
of rainfall increases (Figure 1, Table I [page 213]) yields decrease-in fact,
for all types of grain. Rainfall around the year 1880 was very heavy and the
yield very small, while during the sixties and at the beginning of the
seventies rainfall was low and, therefore, the yield was high: one curve is the
exact mirror image of the other. In addition, variations are quite pronounced;
during the wet lustrum of 1881-85 the crop of wheat was down compared to
the dry lustrum of 1861-65 by 225 kg per hectare or 16%, and that of rye by
even 325 kg or 26%. Even if one combines several lustra, the influence is
still evident. Throughout the entire wet period of 1876-93 10% less wheat
and 19% less rye per hectare was harvested than in the dry period of
1859-75. Consider the meaning of this-over a period of 15 years a loss of
10 respectively ·19 percent! Exactly the same results tum up when the
average harvest is estimated. 407
In view of this obvious influence of climate variations on the yield, it
comes as no surprise that grain prices suffer (compare Figure 2 and 4, Table
II, III, and IV [on pages 214-216]). The development of prices does not
show the same parallel pattern as is evident between the changes in rainfall
and crop yields. There are times when suddenly a country's political
situation, the competitive influence, changing exchange rates, etc., interfere
with this correlation.
The dry period of 1821 to 1840 has a steady strong impact in the ocean-
regions. In all of them grain prices are considerably lower than what they
had previously been. With the beginning of the wet period in the forties
prices start to rise again; a maximum is reached almost everywhere in
1851-55, at the same time or immediately after the maximum rainfall. This
407 With the exception of the lustrum of 1846/50.
nico.stehr@zu.de
228 EDUARD BRUCKNER
Rainfall
1848 53 58 63 68 73
/
78
-
83
"'-
88 91
Wheat harvest ""I," """.
/
100... V
"
%
Rye harvest
%
~
.... 1"""'" 1"""00...
- "' .....
./
"'" "''" "- J1III"
Wheat harvest
kg pro ha ~
..,,-
--- " "' - "-
~
"
Rye harvest ./
" "' "
kg perha ",,-
"" ~
1848 53 58 63 68 73 78 83 88 91
Figure 6.1. Fluctuations of Rainfall and Grain Crop in Prussia. Regenfall = Rainfall;
Weizenemte =Wheat Crop; Roggenemte =Rye Crop.
development is quite evident in all countries of Western Europe, except to a
relatively lesser degree in England. In all of these countries prices then drop
slightly in accordance with less rainfall until 1861-65, Austria being the
only exception. Again this period is followed by a price increase as rainfall
increases as well. It is, however, of short duration and the price already
peaks in the years 1871 to 1875, that is 5 to 10 years prior to the maximum
amount of rainfall, only to fall considerably after. These prices drop despite
the fact that crops in the producer country continue to worsen and do not
begin to improve until the second half of the eighties (compare with Prussia).
Consequently, it has nothing to do with the outcome of the crop at home and
is the result of massive grain imports from America and Russia.
In order to determine the extent of the price changes due to climate
variations, it is advisable to ignore absolute values and to simply deal with
the maximum lustra expressed in percentages of the minimum lustra. This
has been done in the following small table, i.e., for those lustra showing
maxima and minima in the majority of countries.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 229
Wheat Wheat Wheat Rye
England France Belgium Denmark
1816/20 to 1831/35 117 113 115
1851155 to 1831135 111 112 115 118
1851/55 to 1861/65 112 III 112 112
Average 113 112 114 115
Wheat Wheat Wheat Wheat
Prussia Miinchen Zurich Austria
1816/20 to 1831/35 115 118 115 115
1851155 to 1831135 116 117 112 115
1851/55 to 1861165 111 112
Average 114 117 113 115
On the average the price during the wettest lustra is higher by l3% than
during the driest ones - certainly this is a substantial amount.
1808 13 18 23 28 33 38 43 48 53 58 63
L
-"
~
Austria ~
Wheat price /
/
..... ~
Austria ,/ ~ l / ..........
Rainfall
" /
/
" ~
- "-
if'
""".......,. ./
Prussia
Wheat price "
-
",,-
..........
/
~
....
Prussia " ""/ , ~
".
"
Rye price
./ ..........
Prussia / I\..
"- "
Rainfall ./
~ 1/ ~
....
1808 13 18 23 28 33 38 43 48 53 58 63
Figure 6.2. Fluctuations of Rainfall and Grain Prices in Austria and Prussia
The situation as we have just described it for West and Central Europe is
quite the opposite in countries with a continental climate as far as the
correlation between climate variations on the one hand and the harvest and
nico.stehr@zu.de
230 EDUARD BRUCKNER
grain prices on the other hand is concerned. There is a very strong influence
of climate variability on the wheat crop in America (Ohio, Figure 3, Table V
[page 217]). In this case, a decrease in rainfall coincides most visibly with
lower crop yields up to 1861-65; after that both increase. In the wet lustrum
of 1876-80 the crop per acre was a full 50% higher than in the dryer lustrum
of 1861--65. After 1880 a small reduction in the yields reflects the slightly
lower rainfall. In general, grain prices react accordingly all over the United
States. In the sixties (1862-1870) during years with small amounts of
rain-i.e., 3% below the average-the price for a bushel of wheat was 143
cents, in the next 5 years with a rainfall of 6% above the medium-only 100
cents. 40S) However, I would put less weight on price development and more
on the yields because at the same time the introduction of farm machinery
reduced production costs in America enormously.
The situation is similar in Russia. Unfortunately this cannot be verified as
clearly as for the United States because in Russia the official crop statistics
were not kept until 1883. For earlier years, only sporadic vague estimates for
certain areas are available, which are not conclusive for the situation in the
whole empire. However it is a fact that Russia had a number of excellent
harvests in the seventies and eighties up to and including 1888 which
coincides with the last period of high precipitation (see Table V [on page
217]). Although bad harvests did occur, such as the crops of 1879 and 1885,
these were exceptions. However, after 1888 crop yields decreased
tremendously as official statistics show.
Prices are the only indicator for those earlier years. In Moscow (Table 5)
rye prices were high during the dry lustrum in 1831-35, obviously because
yields were low due to droughts. The price fell and remained low in the
following wet period of 1836-55, then rose again. Another drop as was to be
expected from the good yields in the wet period of 1871-85 did not take
place. Undoubtedly, this is related to the extreme devaluation of the Rouble.
Despite the relatively low grain prices abroad, Russian traders still earned a
lot of money, which had consequences for the price at home. 409
40S Calculated according to Francke's figures in Zeitschrift des preuBischen statistischen
Bureaus, 1887, p. 125. The rain is the mean of four groups averages, given in
Klimaschwankungen [Climate Change; Chapter 4b in this collection]. without the Atlantic
states.
409 Exchange rate fluctuations are also the cause that Hungarian grain prices do not reflect any
influence from climatic fluctuations.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 231
1853 58 63 68 73 78 83 88
Russia ./
Increase of "-
Rye Export "- f "
"-
"
I
/
...... ""
Russia
L
~
/
"'"""
Rainfall ~
'\.. II" J
Ohio
".........." /
I
~
/
Wheat Harvest 1
Ohio " "-" .., ",
If
/ ----
"
Rainfall V
"'- JIll'
1853 58 63 68 73 78 83 88
Figure 6.3. Rainfall fluctuations and increase of rye export in Russia and of wheat crop in
Ohio
Let us take another look at climate variability and grain prices in the past
century (Table II [page 214], Figure 4). The interrelations are the same as in
the current century, as far as observations are available. In England rainfall
and wheat prices show exactly the same development from 1701 to 1790.
There is a slightly lower congruency between wheat prices and the harvest
dates for grapes in France as shown in Table II and during the first half of
the century in Germany. This is not surprising since both products may be to
a great extent affected by the amount of rainfall, but not exclusively. A
general trend is obvious though. For example, during the wet period of 1770
prices rise significantly everywhere, which is superseded as well as followed
by periods of low prices and less rain. The years 1710 and 1730 in England,
and 1715 and 1740 in Germany stand out as times when costs of living were
high. This time of high prices at the beginning of the forties does coincide
with war-times in Germany; but since at the same time heavy rainfalls were
recorded, I am inclined to attribute the high costs of living, at least in part, to
the latter. A strong increase in prices and their continuous high level can be
observed everywhere from 1790 to 1820. In Germany this development
could, at least in part, be the result of the climate (compare Figure 4). Prices
again are not noticeably higher than later in the forties. In England, however,
rainfall does not show maximum amounts at this time but some time later.
nico.stehr@zu.de
232 EDUARD BRUCKNER
As a consequence, it may be assumed that the high costs of living are mainly
the result of political tension: All of Europe was heavily armed. 410
17ru III IJ .1 D .a U oM "J ... n 51 6) .. n 71 IJ U .) " I... U II n _ u 3.. 4I} .. Sl !I* &.J &of.
"".....:
Raulll.a
E.....
Ralllfall
~~"'"
"'"""'"
w""''''"
'-
WhaoPn<.
w~ : ......
0. .. ,
o.lr:ol 1M;
""""' "
Rlkinl.u
170J. U I. 2) 1II n JI ,U '" !1 SI 61 .. 11 " IJ .. .) .. ,.. • u " U lit JJ ,. .) '" n )I 6) "
Figure 6.4. Fluctuations of rainfall and grain prices in England, Southern Gennany, and
Switzerland.
III.
If grain crops and grain prices are influenced by the amount of rainfall, it
should also be possible to link climate variations to trade and trade policies. I
do not have all the data necessary to verify the correlation for all countries, a
task that would exceed the purpose of this essay anyway. Nevertheless, I
would like to show in what way climate variability. reflects directly on trade
policy by citing the example of the grain trade between Russia and Western
Europe. 411
Primarily two forces rule the export of grain from Russia to Central and
Western Europe: the demand in the West, and the surplus in the East. All
other factors such as the construction of railroads, improvements in agricul-
ture, trade policies in general can never be considered as prime causes; they
merely serve the purpose of responding to demand or of disposing surplus.
We must expect higher exports if the demand increases in the West and the
East has a large surplus, i.e., when crops are bad in the West due to a lot of
precipitation and for the same reason crops are good in the East. Lower
exports must be expected if in the West demand goes down because dry
410 According to Geering when reviewing my essay about Russia's future as a grain supplier
in Neue Zurcher Zeitung. January 17, 1895.
411 I will mainly use the data from Laves, Getreideproduktion und -handel im Europiiischen
Russland [Grain production and trade in European Russia], Jahrbuch fUr Gesetzgebung. .
2c (Schmo\ler) N.F. V, p. 293.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 233
weather produces good crops and in the East the surplus is reduced because
of bad harvests.
It is quite interesting to be able to recognise this phenomenon clearly in
the development of Russian grain exports. At the beginning of our century,
during the period of high precipitation, which lasted in some countries up to
1820 in others up to 1815, England and Central Europe had crop failures,
and Russia exported relatively large amounts of grain, i.e., from 1800 to
1813, 4 million hectolitres annually. It is during this time that rapid
settlement in Southern Russia's steppe takes place. As early as 1816 the port
of Odessa exported 7 to 8 million metric hundredweight, an amount that was
not reached again until 1878 and surpassed in 1879. The West now begins to
enjoy good harvests during the dry period. England prohibits the import of
wheat due to its own surplus. Throughout central and western Europe grain
prices are low, demand is low because home production is self-sufficient,
and by 1825 to 1840 Russia's export goes down to 3,4 million hectolitres
annually because of drought. Again another moist period follows. England
abolishes the grain tax in 1846, bad harvests increase the price of grain.
Transportation and certain laws are improved which benefit grain imports in
Western Europe, because demand increases, and success is not far behind: in
the years 1844-53. Russia exports 11 112 million hectolitres grain annually,
almost four times as much as in the thirties. Export decreased during the
Crimean Wars, but even in the years 1856-64 it was slightly below the
amount of 1844-53. This backdrop was explained by the fact that so many
people lost their lives during the Crimean Wars. However, how much do
100,000-200,000 people count, and the number of dead was certainly lower,
compared to a population of more than 50 million of which 12 million were
men in the work force? Crop results seem to give a much more plausible
explanation for the decrease of the export: in the West good harvests are the
result of fairly dry weather, while in Russia the same weather produces bad
harvests. Since the middle of the sixties Russian export experiences an
unprecedented big increase coinciding with wet weather conditions, during
which harvests in Russia are extremely good, but in Western Europe
extremely bad. As a result Russian grain exports clearly reflect the
variability of the climate.
To convert this more precisely into numbers, the columns 7, 8, and 9
were created in Table V [page 217]. The upward trend of the grain export
continues due to a steadily growing population in the West as well as to
improvements in transportation. It is, however, briefly interrupted or at least
slowed down because of climate variations. In order to isolate the climate's
influence, the general upward trend had to be eliminated. I achieved this by
drawing a line through all the export figures of each lustrum according to the
method of least squares. For each lustrum I then determined the differences
nico.stehr@zu.de
234 EDUARD BRUCKNER
between the figures gained through this method and the actual export
figures. Whenever these differences were positive, the export's increase had
accelerated; were they negative, the increase had decelerated. The figures of
column nine in Table V relating to the rye export were arrived at in this way.
These figures were adjusted and graphically depicted as shown in Figure 3.
A comparison with the variations in the amount of rainfall clearly confirms
that the increase of grain export is accelerated in wet periods, while it slows
down in dry ones, even a slight decrease takes place. Since this correlation
has been verified for the entire 19th century, chance is obviously out of the
question: climate variability has a clear impact on the grain export of Russia.
This does not mean that we deny any influence by various other factors such
as competition, political constellations, customs and tax regulations, etc. It
seems, however, that none of these factors can fully hide the influence of
climate variations.
Finally, a further question comes to mind: What will weather conditions
in the next decades be like as a result of climate variability and how will they
influence trade?
From the year 1000 on, not less than 25 full climate variations have been
identified and consequently there can be no doubt that the 26th will not fail
to occur. All indications are that the last wet period peaked in the year 1880;
from then on or even at least in part from the year 1885 on, rainfall
decreased significantly in all parts of the globe. Our tables offer examples.
Lately Heinz mentioned this reduction in rainfall for Russia from 1881-85 to
1890.412 It has progressed further since then. I predicted this development in
the year 1888 and mentioned at the time 413 that we are approaching a dry
period, which can be expected to peak around the turn of the century. I
added furthermore that the coming dry period will create major economical
crises in the continental regions in particular, and will destroy the livelihoods
of thousands, if not hundreds of thousands. Events have already confirmed
this. The Russian crop of 1888 which was not exceptional, yet might still be
regarded as good, was followed by the bad harvest of 1889, the mediocre
one of 1890, and finally by the complete crop failure of 1891. The harvest of
1892 was again quite mediocre. In 1893, however, Russia was blessed with a
good harvest. 414 This should not be surprising. For, periods of dryness are not
characterised by a series of bad harvests but rather by their increased
frequency and a decrease in the frequency of good harvests. In short,
412 Repertorium fUr Meteorologie XVII, No.2.
413 In my inauguration lecture in the small auditorium of the University of Bern in May 1888.
Compare also Verhandlungen des VIII Deutschen Geographentages zu Berlin, Berlin
1889.
414 According to the official harvest reports of the Central Committee for Statistics of the
Russian Ministry of the Interior.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 235
indications are that Russia is on the brink of a period with harvests of
inferior quality. This fact alone should have an adverse effect on export. It is,
of course, quite certain that Russia alone would be in a position to satisfy all
of Europe's demand for grain, even in bad years, simply by intensifying her
agricultural methods. However, in times of crop failures, agriculture cannot
be expected to expand; because that would require capital, and capital
especially will be scarce and in part be used up as a result of bad harvests:
agricultural decline is the consequence. Russia's crop failure of 1891, for
instance, caused a noticeable reduction of agriculturalland. 415 But a reduced
production would not put too severe a restriction on Russian exports if there
was not a second more important aspect involved: The need for Russia's
grain is today no longer as urgent in the West as it was 5 or 10 years ago. Its
demand has decreased, which again is a result of climate variations. For, in
the West rainfall has decreased over a number of years as well, and
consequently higher yields were achieved. The year of 1893 was so dry that
agriculture suffered damages in certain areas, hay production in particular.
Nevertheless this is the exception; generally agriculture is expected to
benefit from a dry period.
It is quite remarkable that the director of trade statistics in the Swiss
Department of the Exterior, T. Geering, emphasised that Europe's current
trade policy and in particular the grain taxes of Germany and France are in
complete accordance with the theory of climate variability.416 "Grain prices
have fallen as low as never before. The demand (of the West) for imports is
relatively small because production at home is sufficient at an unusually high
degree and consequently it will be more likely that grain imports from
Russia and North America to Central and Western Europe will decline. The
tax policy of the continental Great Powers (i.e., France and Germany) should
therefore, irrespective of other economical considerations, continue to get
support from the prevailing climate as a factor in production over the next 10
years to come."
During the next wet period however which is expected to peak in about
20-25 years, a tremendous increase of grain imports from those countries
with harvests SUbjected to continental climate variations and the price
increases in the West may perhaps reverse trade policies in a free-trade
direction as it happened in England in 1846 as a result of bad harvests.
415 According to the official pUblication Resultats generaux de la recolte en Russie 1892
(St. Petersburg 1893), the agricultural land in the European part of Russia had diminished
by 927,000 hectares in that year compared with its size in 1887, when the last estimates
were made. This reduction is very extensive in areas hit by the crop failure of 1891,
whereas the western and northwestern districts show increases.
416 Geering
nico.stehr@zu.de
236 EDUARD BRUCKNER
There is a peculiar kind of interaction between countries in which har-
vests are exposed to either the oceanic or continental type of climate varia-
tions: a relationship of mutual compensation regarding grain production. If
one country has a good harvest, the other has a bad one and vice versa. This
definitely must lead to prosperity as has been indicated in the figures for the
seventies and eighties in our Tables III [page 215] and IV [page 216]. Long-
term variations in grain prices will at first be subdued in the wet regions of
Western Europe by the fact that price increases do no longer occur during
wet periods. Prices have always been low in dry periods as a result of good
harvests at home; they now will also be low during wet periods because of
the rich crops of the big grain suppliers Russia and America as well as India.
The costs of this development must be carried by the agricultural industry of
the West which will have to go through critical times because each period of
wet weather will produce poor harvests at home and an oversupply of cheap
grain from countries enjoying good harvests at that time, as it happened in
the eighties and in part still happens today. The less self-sufficient the West
will be in satisfying its own demand for bread and the bigger the surplus
produced by America and Russia is, forcing them to sell at any price, the
closer the West will follow the pattern of price variations in these large grain
producing countries. Most· likely, this trend in England will be noticeable in
the very near future, while Germany is still basically self-sufficient in bread
supply.
Let us end here!
Up to now it was considered, as we emphasised earlier, that those factors
beyond human power which influence harvest losses are mere chance occur-
rences, "to such an extent that they were completely ignored in the practical
considerations concerning agricultural production and tax policies, and it
was assumed instead that chance would take care of them".417 The human
mind was regarded as the sole factor in economical development. Trade
policies, social conditions, the relentless fight for survival, these were the
only factors taken into consideration when economical phenomena needed
explaining. I am far from underestimating the enormous force of the human
mind in this matter. But one factual influence of tremendous importance in
the form of climate variations, that is the change between favourable and un-
favourable weather conditions, plays an additional role that can no longer be
overlooked. Climate variations influence crops and grain prices. I feel we
have proven this here. That the entire economy and the course of history is
affected is rather obvious. How much more important this influence is as op-
posed to those factors created by the human mind, I do not dare decide. It
has so far not been completely dismissed while trade policy has repeatedly,
if unintentionally, taken climate variability into account.
417 Geering as mentioned before.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 237
APPENDIX: TABLES AND SOURCES
The method of tabulation has been explained above on page 202. Extremes
are marked in bold or with an asterisk. My essay Climate Variations since
1700' is initialised as "C. V." when cited.
Table 1. Variations of Crop Yields in Prussia!
Rainfall~eid per Average-Harvest> (%)--. Yield per Hectare 4 (kg)
(%) Wheat Rye Barley Oats Wheat Rye
1846/50 2 7.4 9.6 4.0 -2.0
1851/55 3 -2.2 -3.8 2.6 1.8
1856/60 -9*. 2.4* 6.6* -9.8 -9.0 1385 1218
1861/65 -7 2.0 2.6 7.0* 8.2* 1399 1250*
1866170 6 -2.0 -2.2 -2.6 -3.4 1410 1220
1871175 -7 1.6 -1.0 -0.6 -1.8 1519* 1249
1876/80 10 -11.0 -13.8 -9.4 -8.4 1360 1066
1881/85 5 -18.6 -23.2 -19.2 -21.6 1175 926
1886/90 3 12.6 -21.6 -17.6 -11.6 1264 947
1891/93 -2 -5.0 -14 -17.3 -19.3 1367 1038
DryS -4.2 1.0 1.6 -1.5 1428 1234
Wet6 4.0 -11.8 -18.2 -15.2 1291 994
Addendum and notes to Table I:
1 Old State, i.e., the 8 old Provinces.
2 Average of 23 Measuring Stations in North and Central Germany in C.V., p. 158,
updated. These unpublished results from observations of 1891-1893 were kindly given
to me by the Koniglich preufiisches meteorologisches Institut.
3 Up to 1877, according to estimates from the annual harvest table that were taken prior to
the actual harvest by estimating the crop in the fields. These estimates are known to be
too low (compare also A. Kremp in Hildebrand's lahrbuch N. F. IX, p. 358). Based on
the data for 1859-76, Engel has defined the term "average harvest" (Miltelernte) by
using the yields (in bushel per acre) from the Erdruschtable (Zeitschrift des preuBischen
statistischen Bureaus, 1878 p. 401). By setting the average of the estimates from the
harvest table of 1859-76 equal to 100, I determined the average deviation by which these
estimates are too low, namely the estimates of the grape harvest by 12%, rye by 16%,
barley by 13%, and oats by 10%. I increased the estimates accordingly, then expressed
the lustra averages in deviations from 100. Although these reductions distort the resulting
figures somewhat with regard to climate variations, harvests vary by much more than
those reduction percentages, therefore even non-reduced figures will still reflect their
variations. In 1877 these estimates end. From 1878 on the amount of harvested grain per
hectare (ha) is included in the Prussian statistics (results of the crop yields). These
numbers were put in percentages of Engel's average value and the columns were
continued until 1893. For rye, only winter rye was included because summer rye is
hardly ever grown, but for wheat and barley the average yield from summer and winter
crop combined was added. Some of the average values 'for the 'old' [Prussian] state had
to be determined anew with East and West Prussia combined as one province as it used
to be before.
• [See Chapter 4 of this collection].
nico.stehr@zu.de
238 EDUARD BRUCKNER
4 Up to 1877 according to the Erdrusch tables, which up to 1866 are published by Meitzen
(Boden und landwirtschaftliche Verhandlungen des preuBischen Staates III), and until
1870 by Engel (Zeitschrift des preuBischen statistischen Bureaus). For the years 1871-77
I obtained data from hand-written notes of the koniglichen preuBischen statistischen
Bureaus. From 1878 on crop figures for grain are included in the Prussian statistics. The
years from 1856 to 1858 are missing
5 1856-75.
6 1876-93.
Table II. Variations of Grain Prices 1701-1800
England France Germany Austria
Rain- Rain- Date of Miinchen Berlin
fall Wheat fall Wheat Grape Wheat Wheat Wheat
{%}l Price2 {%}3 Price4 Harvest 5 Price 6 Price? Price~
1701/05 0 28.6 5.2 17.6 31*
1706/10 0 42.1 -1.0* 12.6* 33
1711/15 11 40.9 5.4 18.1 42
1716/20 -6 34.0 2.6 15.5 42
1721/25 -6* 31.0* 4.6 14.7 34
1726/30 3 41.4 -1.2* 15.9 32*
1731/35 -6 27.1 4.6 13.4*
1736/40 0 36.1 2.4 18.4
1741/45 -12* 29.0* 6.8 24.3
1746/50 -3 29.8 1.7 18.2
1751/55 6 33.6 5.8 14.9
1756/60 4 42.0 -{J.7 11.1 0.5 19.9
1761/65 5 34.8 -3.3* 10.1* -2.3* 18.8
1766/70 9 51.1 6.5 15.5 7.8 22.7 50
1771/75 10 51.5 0.3 16.2 2.3 32.3 63
1776/80 -6* 40.2* -{J.7 13.4 -1.3 17.6* 40*
1781/85 -2 48.7 -7.2 14.8 -4.0 18.5 48
1786/90 -2 47.2 0.5 17.2 0.1 22.3 54 2.9
1791/96 -2 53.6 -5.7* 23.9 56 3.0
1796/00 -3 73.4 0.7 30.5 64 3.3
Dry9 2.1 15.1 32
Wet 10 4 37.2 4.2 16.1 40
Dry 11 -6 32.5 1.9 15.9 32
Wet l2 -1 34.9 4.8 19.1
Dry 13 -8 29.4 -2 10.6 -{J.9 19.3
Wet l4 7 42.6 3 15.8 5.0 27.5 56
Dryl5 -3 52.6 -4 14.1 -2.8 20.6 50
Addendum and notes to Table II:
Up to 1730 the averages are from several West European stations and later only from
England. C.V., pages 188 and 190.
2 Up to 1770 prices in Shillings (with decimal fraction) per Winchester Quarter at Eton,
average per year. Later for all of England-official average per year and Imperial
Quarters. Tooke and Newmarch, Geschichte und Bestimmung der Preise [History and
Definition ofPrices], translated by Asher. Dresden ,1859, Vol. II, p. 512.
3 The Dates for Harvesting Grapes in Central France (C. V., p. 264); Minus (-) is too early,
meaning warm and dry. Numbers stand for days.
4 Francs per hectolitre. Statistische Monatsschrift, 1877, p. 357.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 239
5 In Southwest Germany and Switzerland. C.V., p. 225.
6 Mark per Bavarian bushel. Zeitschrift des preuBischen statistischen Bureaus, 1886,
p.228.
7 Silver Penny per Berlin bushel, ibid., p. 225.
8 Austrian Guilders. Tafeln zur Statistik des Steuerwesens 1858, Wien
9 For Germany 1701-10.
10 For England 1701-15, Germany 1711-25.
11 England 1716-25, Germany 1726--40.
12 England 1726--40, Germany 1740-55.
13 England 1741-50, France and Germany 1756-65.
14 England 1751-75, France and Germany 1766-75.
15 England and France 1776-1800, Germany 1776-95.
Table III. Variations of Grain Prices in Western EuroQe 1801-85
England France Belgium Kebenhavn
Rain- Wheat Rain- Wheat Rain- Wheat Rain- Rye
fall {%ll Qrice 2 fall {%13 Qrice4 fall {%15 Qrice 6 fall {%17 Qrice8
1801105 -12* 80.2 0 21.8 16 8.6
1806/10 -3 87.9 4 18.1 12 8.3
1811115 0 94.3 1 23.9 -12
1816/20 -2 80.8 -3 25.3 5 25.6 4
1821125 8 57.3 -6 16.6* 1 13.2 -2 2.3*
1826/30 1 61.6 1 20.2 3 18.9 4 4.4
1831135 -2* 52.7* -8* 18.2* -10* 17.4* --4 4.0
1836/40 -2 61.2 1 19.9 -5 19.2 -10* 4.0
1841145 3 54.8 9 19.6 6 19.9 8 4.7
1846/50 3 51.9 -1 19.9 -3 20.1 -3 4.6
1851/55 -3 55.9 3 22.5 I 25.3 I 7.3
1856/60 -1 53.3 4 21.8 -9* 22.8 -10* 5.4*
1861165 -11* 47.5* -8* 20.3* -9 21.3* 1 5.9
1866170 --4 54.6 -6 22.7 10 24.1 -3 7.0
1871175 1 54.7 -1 23.8 -2 12 9.5
1876/80 18 47.5 11 22.4 12 -3 8.7
1881185 6 40.1 * -1 19.5 10 0 7.4
Wet 9 -8 84.0 22.3
Wet 10 2 72.0 1 18.3 3 19.2 14 8.4
Dry II -2 57. --4 20.7 -8 18.3 -3 3.7
Wet l2 3 53.4 3 22.3 1 21.8 2 5.5
Dry 13 -5 52.8 -5 20.9 -9 22.0 --4 6.1
Wet l4 8 47.4 5 10 24.1 3 8.5
Addendum and notes to Table III:
See note 1 in Table II.
2 In Shillings and decimal fractions per Imperial Quarter. Up to 1855 according to Tooke,
later according to Francke in the Zeitschrift des preuBischen statistischen Bureaus, 1887
p.124.
3 Mean value of the averages of three groups each (without Mediterranean France) for
France in C.V. on p. 166.
4 French Francs per hectolitre. Up to 1870 according to "tableaux des prix moyens
mensuels et annuels de l'hectolitre de froment en France 1800-1870". Paris 1872. Later,
according to Francke, op. cit.
nico.stehr@zu.de
240 EDUARD BRUCKNER
5 Up to 1830 averages from Northern France and Holland, later from Holland and
Belgium, C.V., p. 167.
6 Fr. Francs per hectolitre. Statistische Monatsschrift III, p. 396. Figures have been
adjusted to a single unit.
7 Averages from K0benhavn and Lund, C.V., p. 161.
8 Price in Rigsdaler Silver (rounded to one Decimal) per ninde = 1.3912 (hI) hectolitres.
Up to 1870. Statistische Monatsschrift III, p. 397. Later in Francke, Zeitschrift des
koniglichen preuBischen statistischen Bureaus, 1887, p. 124. 1871 and 1872 are missing;
for the lustrum 1871175 which is not available, the years 1873-77 were used as
substitute.
9 Refers to England in 1801-10.
10 Refers to England in 1811-30, France 1801-20, Belgium 1816-30, K0benhavn 1801-10.
11 England 1831-40, France 1821-31, Belgium 1831-40, K0benhavn 1821-40.
12 England 1841-50, France 1836-60, Belgium and K0benhavn 1841-55.
13 England 1851-70, France 1861-75, Belgium 1856-65, K0benhavn 1856-70.
14 England and K0benhavn 1871-85, France 1876-85, Belgium 1866-85.
Table IV. Variations of Grain Prices in Central Europe 1801-1885
Southern Germany and
Prussia Switzerland Austria
Miinchen Ziirich
Year Rain- Rain- Rain-
fall l Wheat Rye faU 4 Wheat Wheat fale Wheat
{%} Prices 2 Prices3 {%} Prices 5 Price6 {%} Prices8
1801/05 5 44.9 5.1
1806/10 12 33.0 -3 4.4
1811/15 12 33.7 16.2 9 4.3
1816/20 -3 206 152 -8 45.9 20.4 1 4.7
1821125 -10* 113* 76* -15* 21.0* 10.5* -6 2.8
1826/30 -3 131 98 -2 22.4 11.6 -3 2.7*
1831/35 -13* 134 103 -10 24.8 13.3 -14* 3.1
1836/40 -1 143 99 -1 22.6 12.3 -1 2.8
1841145 3 154 114 7 29.3 14.3 2 2.9
1846/50 2 181 131 -1 33.0 15.4 9 4.1
1851/55 3 214 177 2 41.3 16.3 2 4.6
1856/60 -9* 209 154 -11 14.6 -3 5.2
1861/65 -7 188* 138* -17* 14.0* -14* 6.1
1866170 6 220 172 3 15.8 2 6.6
1871175 -7 235 179 -5 17.8 -3 6.6
1876/80 10 211 166 12 14.8 22 5.9
1881/85 5 190 160 2 13.5 10 5.2
Wet9 -3 206 152 5 39.4 18.3 2 4.6
DrylO -7 130 94 -7 22.7 11.9 -6 2.8
Wet II 3 183 141 3 34.5 15.3 4 3.9
Dry 12 -8 198 146 -14 14.3 -8 5.6
Wet 13 4 214 169 3 15.5 8 6.1
Addendum and notes to Table IV:
1 Averages of the 23 stations in Northern and Central Germany; see notes 2 of Table I.
Prior to 1831 averages of the Central European stations, C. V., p. 158.
2 Annual average for the Prussian State in Mark per 1000 kg. Francke in Zeitschrift des
koniglichen preuBischen statistischen Bureaus, 1887, p. 121.
nico.stehr@zu.de
INFLUENCE OF CLIMATE VARIABILITY ON HARVEST 241
3 ibid.
4 Average according to C.V., p. 167, but from 1871 on without Mannheim, Karlsruhe,
Zurich, and Aarau, because observations at these locations were no longer comparable
with the older ones (comp. SchultheiB in lahresbericht des Centralbureaus flir
Meteorologie und Hydrographie im GroBherzogtum Baden, 1888 p. 57 a. 1890 p.75).
5 Mark per Bavarian bushel. See above note 6 of Table II
6 Fr. Francs per 50 kg. Statistische Monatsschrift, III p. 365. From 1871 on Zeitschrift flir
schweizerische Statistik 1883 and 1887.
7 Average for Bohemia and Austrian Alpine regions, C.V., p. 167.
8 Austrian Guilders per peck, Statistische Monatsschrift, III, p.365. From 1871 by Francke
op. cit. Converted under the assumption that one peck [Metze=3.44litre] equals 47 kg.
9 1801-20.
10 1821--40.
11 1841-55.
12 1856-65.
13 1866-85.
Table V. Variations of rye prices and rye export in Russia and yields in America
Russia United States
Rye Export Wheat
Rainfall l Rye Price in 100 2000 Hectolitres Rainfall Bushel
Year {%} Mosco~ Observ. Ca1cul. Diff. {%} p. Acre 5
1801/05 38
1806/10 29
1811115 31
1816/20 36
1821/25 35
1826/30 26
1831135 -18* 42*
1836/40 2 39
1841/45 8 39
1846/50 -2 37 15
1851155 6 38 21.4 13.1 8.3 -2 13.2
1856/60 -8 51 34.2 37.7 -3.5 2 12.5
1861/65 -11* 30.2 62.3 -32.1 -5* 10.7*
1866170 1 } 63* 46.2 86.9 --40.7* -1 12.1
1871175 2 70 124.0 111.5 12.5 -5 13.7
1876/80 12 79 193.4 136.1 57.3 10 15.3
1881/85 8 90 141.0 160.7 -19.7 4 15.0
1886/90 2 188.4 185.3 3.1
Dry6 -18 42
Wee 4 38 8.3 0 12.8
Dry8 -6 57 -24.5 --4 12.2
Wet9 6 80 13.3 7 15.0
Addendum and notes to Table V:
Mean value of the 3-group averages: NW-, SW-, and SE-Russia as in C.V. on p. 167,
completed up to 1890.
2 Kopecks per Pud. According to Annuaire statistique de la Russie [Annual statistics of
Russia], 1890, p. 131. The years 1863-69 are missing.
nico.stehr@zu.de
242 EDUARD BRUCKNER
3 Iuraschek, Ubersichten der Weltwirtschaften 1885/89 [Summaries on World Economics
1885-89] p. 44. 1890 according to the Statistical Abstract for the principal and other
foreigne countries in each year from 1881 to 1890/91.
4 Group B.S. Interior, East, C.V., p. 168.
5 In Ohio, C. V., p. 277.
6 For Russia 1831-35.
7 Russia 1871-55, Ohio 1851~0.
8 Russia 1856-70, Ohio 1861-75.
9 Since 1871 respectively 1876.
nico.stehr@zu.de
Chapter 7
Weather Prophets *
The desire to look into the future is deep-rooted in mankind. Hence the fact that
fortune-telling is so enormously popular, having more followers even in our
time than one would be inclined to assume. Those who consider themselves
educated do of course no longer revert to palmistry nor do they consult the
cards, and if so, only in secrecy: They are ashamed of an action incompatible
with their level of education. There is only one category of fortune-telling that
flourishes openly even today, which has its prophets as well as its believers
even among the highly educated: it is the foretelling of the weather. Even
respectable newspapers from time to time print predictions from weather
prophets. Besides Falb's weather prognoses, Swiss newspapers carry weather
prognoses by local prophets as well, such as by the high school teacher
C. Marti in Nidau, the engineer Gladbach in Aarau, and Jules Capre in Chillon
for the French-language newspapers. The public reads these prognoses, follows
them up, and finds that in some cases the predicted weather actually occurs.
Such correct forecasts win followers over and they gather around their prophets
in considerable number. Is the belief of this group justified? We will try to
answer this question.
Weather predictions have been around since the earliest documented times,
their oldest versions being based on religious convictions.418
In antiquity, Greek and Roman gods controlled the weather in an absolutist
manner, headed by Zeus, father of the Olympus. During a thunderstorm,
several gods got into the action at the same time: Zeus through his thunderbolts,
Aeolos sent out the winds under his command, and Iris stretched the rainbow
across the sky. The ancient Germanic tribes did not think much differently, and
even today we find similar views amongst the beliefs of many peoples. The
North American native Indian hears the voice of the 'Great Spirit' in the
• Wetterpropheten, lahresbericht der Berner Geographischen Gesellschaft, 1886.
418Compare C. Lang. Weather prophesies in old and modern times, publ. in Samm1er,
Addition to the Augsburger Abendzeitung; 1889/90.
243
nico.stehr@zu.de
244 EDUARD BROCKNER
thunder; and the Caffre [member of the Bantu-tribe] will say when it thunders:
"The old man is bowling419", in reference to someone living beyond the clouds.
The role of mediator between humans and gods fell to the priests; conse-
quently they were the ones who had to consult the gods about future weather
outlooks and to announce the answers. Apollo for instance had established
some kind of information bureau in Delphi where all kinds of predictions, in-
cluding weather forecasts, were handed out in exchange for money and kind
words. In many cultures, priests were not· only weather prophets but also
weather makers, using symbolic acts, prayers, sacrifices to the gods, etc., in
order to persuade or even to force the gods to send a certain weather system. If
the friendly gods did not respond, the evil ones were called upon-in medieval
times, even the devil himself Members of the weaker sex in particular were
often suspected of having influenced or bewitched the weather in cohorts with
the devil to the detriment of their kinsmen; many proclaimed weather-witches
were burned at the stake. The belief in witchcraft was seemingly justified as
some voluntarily claimed to be witches and have carried out symbolic acts with
the intention to exert witchcraft; in such cases their confession came from a bad
conscience.
In Central and Western Europe, weather prophecies based on religious
grounds are rarely published or believed in today. This is different with weather
prophecies based on astrology. These, too, are ancient: Chaldean, Babylonians,
and Egyptians attempted to interpret future events from the constellation of the
stars. Greek natural scientists of the rank of Aristotle, Hypocrates, and Strabo
engaged in this art. Rome paid homage to astrological superstition; even first-
rate minds, such as Cicero, Virgil, or Tacitus were swayed by it. There were
many astrological schools where the skills of fortune-telling were systemati-
cally acquired. This pseudo-science had its prime time in the Middle Ages.
Astrological superstition is based on the so-called geocentric ideology in
which Earth is the centre of the universe. Anything beyond Earth is there only
to serve this planet and has a certain purpose in relation to it. The constellation
of fixed stars is seen as a clock's dial and the planets whose positions in
relation to the fixed stars are constantly changing, as the hands of the clock
pointing out future events, including the weather. As seven planets were known
to exist, it was logical to connect each day of the week with one of them; but
also each year had a specific planet that determined that year's weather. Saturn,
the highest ranking planet, was supposed to be an enemy and destroyer of
nature, malevolent, cold, and wet by nature; Jupiter, rather wet than dry and
warm; Mars hot and rather dry than wet; Venus rather wet than dry-thereby
warm; Mercury cold and dry; and the moon rather wet than cold and dry, at the
same time windy. The sun, in contrast, was to be everyone's friend, not too
cold and not too hot, yet dry. Because each weekday was controlled by a
419 In the original words, 'plays'.
nico.stehr@zu.de
WEATHER PROPHETS 245
particular planet, the influence of the planet reigning throughout the entire year
could be weaker or stronger, etc., on this day. The ruling planet of each year
can be detennined quite easily: you only have to divide the year by 7; the
remaining number refers to the number of the planet that rules the year (1. Sun,
2. Venus, 3. Mercury, 4. Moon, 5. Saturn, 6. Jupiter, 7. Mars); e.g., the year
1902 divided by 7 leaves 5; the ruling planet of that year is Saturn. The year
1902, should therefore, under the reign of this "malevolent destroyer of nature",
tum out to be cold and wet.
How deep-rooted the belief in these astrological prophecies was at the time
is best exemplified by the impression Johann SWffier's prognosis had on his
contemporaries. 420 In his calculations of the planets' future constellations,
this-by the way quite able-astronomer from Tiibingen discovered in 1518
that, in February of 1524, the planets Saturn, Jupiter, and Mars would meet in
the constellation of Pisces. As Saturn and Jupiter were both considered wet
planets, and Mars to be misanthropic, and the sign of Pisces where the
constellation was expected was a priori associated with water, Stoffier, in his
letter to the emperor Charles V, did not hesitate to predict a flood of biblical
proportion for February 1524. Though some scientists contradicted Stoffier, the
general popUlation was in great fear of the prophesied flood. Those who lived
near the sea or next to a river tried to sell their property and moved to higher
grounds. Others built arks to survive the flood as Noah had done. Many were
driven mad with fear. Mayor Hendorf of Wittenberg , according to Luther,
made extensive arrangements in the attic of his house for a survival of the
flood, which included a substantial stock of beer-"in order to at least have
plenty to drink should the deluge come." The dreaded February 1524 came: the
weather in Europe was predominantly bright and nice, there was little rain-no
trace of a flood. One would now assume that such a total failure would cure
mankind of astrological superstition, but far from the truth. There was no
shortage of excuses for the misdiagnosis. The monks declared that their prayers
prevented the flood that otherwise inevitably would have occurred. Biblical
scholars pointed out that God's promise to Noah was overlooked in the
prophecy, namely that there would be no second flood after the biblical one,
which He sealed with a rainbow; that is why the flood did not take place
despite the constellation. Historians however did not at all relate the disastrous
constellation to a meteoric deluge but to a political one-the peasants' uproar
that broke out in 1524!
Astrology continued to thrive happily although its foundation was soon
taken away by the Copernican world view that no longer considered earth the
centre of the universe, but the sun. This reduced the earth to the rank of just
another planet. Even such astronomers as Tycho de Brahe or Kepler could not
escape the prevailing views of their time and included astrological prophecies
420 Compare G. Hellmann, Meteorologische Volksbucher, Himmel und Erde, III, 1891.
nico.stehr@zu.de
246 EDUARD BRUCKNER
in their own prognoses without personally believing in them. "Astrology is the
foolish daughter of astronomy; but she provides for her mother," said Kepler.
Science had to earn its bread and had to take it where it could find it.
Astrological superstition regarding weather developments has been codified
in the hundred-year calendar first published by the abbot Martin Knauer. Even
today, this popular book is being re-edited over and again fmding faithful
buyers. 421 The farming community in particular swears by its "CentenniaL". A
calendar that does not include the forecasts of the "Centennial" will not be
bought. 'The Limping Messenger' ["Hinkender Bote"] of Neuenburg learned
this the hard way: When during one year the astrological forecasts were
omitted, the number of subscribers dropped so dramatically that the publisher
was forced to reintroduce them in next year's edition, bowing to superstition.
Even today all relevant calendars in Switzerland, e.g., 'The Swiss Fanner'
["Der Schweizer Bauer"], the various 'Limping Messengers', include the
predictions of the hundred-year calendar, which, together with the astronomical
data of each year, are edited by qualified Swiss scientists. Kepler's proverbial
words are just as valid today! Even the calendar's producers seem to agree with
the view point expressed by one of their editors, who signed on the front page
of one of the calendar's editions in the 18th century with "Tiehrhawnu", which,
read backwards, means" Unwahrheit"[''untruth''] .422
Recently, weather prophets have emerged who seem to rely on a more
scientific basis turned up lately. Most of them give the moon a major role. A
priori, it seems indeed likely that the moon has a major impact on the weather,
as obviously everyone can see its effect on the ocean along the coasts in the
spectacular phenomenon of the tides. It is mainly the moon's gravity that
makes the ocean level rise and fall twice within 24 hours. A similar, though
weaker, impact is exerted by the sun. If these two effects are combined, as is
the case during full moon and new moon, the tidal movements are particularly
strong. The ocean is in a state of tremendous agitation. It was natural to assume
that the moon had a similar tidal effect on the earth's atmosphere and
consequently on the weather. However, a close scrutiny of air pressure
observations showed no noticeable atmospheric tidal movements. Nonetheless
a number of weather prophets seized on this idea and developed systems of
weather prophesies based on the moon's tidal effect; as Overzier did 20-15
years ago, also Baron Friesenhof, and Gustav Jager, who is more known as a
would-be apostle than a weather prophet, and also RudolfFalb. 423
Falb is the most popular of all, partly perhaps because he invented a slogan
for his weather predictions: his "critical days" are known all over the world. He
421 I have for example: Dr Martin Knauer's centennial calendar for the 19th and 20th century.
Bern, J. Heuberger's Publishers, 1883.
422 According to C. Lang.
423 When this lecture was held, Falb was still alive.
nico.stehr@zu.de
WEATHER PROPHETS 247
calls those days 'critical' during which the tidal factors' effect is particularly
strong. Just as the ocean's tidal movement is strongest under the influence of a
new moon and a full moon, Falb applies this phenomenon to the earth's
atmosphere. His 'critical days' are always days when there is a full moon or a
new moon. Furthermore, just as these tides during full moon or new moon are
particularly strong whenever the moon, or even sun and moon at the same time,
are close to earth, the atmospheric 'tides' are supposedly of equal intensity
during such days. In this manner, Falb differentiates between 'critical days' of
the first, second, and third order.
Faithful believers in Falb's predictions are to be found among both city and
country dwellers; his "weather calendar and index of critical days" that he
publishes every six months is available everywhere. Newspapers pay large
sums for the copyright to print his prognoses. In fact, in many cases these
prognoses prove right; the weather turns out to be just as Falb had predicted it
to be during the 'critical days'. Chance hits like these are exploited by Falb and
dazzle the pUblic. That, nonetheless, his prognoses are worthless and their
fulfilment says nothing in his favour shall be outlined briefly.
Let us first ask which kind of weather conditions Falb considers typical for
the critical days. He writes in his weather calendar of 1901 I: "1. An increase in
barometric minima and low pressure, cyclonic winds and increased precipita-
tion in general. 2. Thunderstorms in winter or at a time of day when they rarely
occur (nights, mornings). 3. Snowfall in the upper mountains in the summer or
in regions where it seldom occurs. 4. Thunderstorm, accompanied by a snow
storm, at the same location. 5. The first thunderstorms in spring and the first
snow fall in autumn. 6. In high altitudes influx of warm southern air masses
saturated with moisture which is indicated either by a sudden thaw or a deep
blue sky and an unusually clear atmosphere, and which then engages in a
struggle with opposed northern air masses characterised by Cirrus clouds or
clouds in general with a tendency to stripe formations," etc., "rainbows, local
showers and frequent changes between rain and sunshine, the so-called April
weather, seem typical conditions." As we can see, it is a whole bouquet of
many different types of weather conditions, and all of them are typical for
Falb's "critical days"!
The critical days are not only in effect on that particular day itself, but may
have an impact two days earlier or even occasionally two or three days later.
Falb considers a total of five to six days to be under the influence of one critical
day. As, according to Falb, each year has 24-25 critical days, more than a third
of the entire year is affected. Falb adds: "That does not mean that the predicted
events have to occur on each of the days described, but that as far as the moon
is a factor, weather conditions tend to be unbalanced on these days." In other
words, the fact that the predicted weather does not materialise cannot be used
nico.stehr@zu.de
248 EDUARD BRUCKNER
against Falb's theory, while on the other hand, correct predictions serve as
proof of his method.
It must be pointed out however, that correct predictions in themselves are
just as useless in proving Falb's theory as incorrect predictions are in
disproving it. The frequency of correct predictions is completely meaningless;
it blends no one but the amateur.
This becomes immediately evident if we ask the following question: Is the
occurrence of an event on a critical day in fact the necessary outcome of this
critical day? Obviously this would only then be the case if this event does not
occur on any other days or at least less frequently than on critical days. An
explicit example will make this clear. Let us suppose someone prophesied that
the sun will rise on Thursdays. He then observes every Thursday and fmds out
that the sun does indeed rise each and every Thursday. He then proclaims: I am
right-Thursday is the sun's critical day for rising. Here again we have a
prognosis and a whole lot of positive returns. Yet, this result is completely
meaningless because the sun rises on all other days as well, not just on
Thursdays. This is the type of mistake Falb makes. To verify his theory, he first
of all needs to identify the frequency of the events which he deems characteris-
tic for critical days for all the other days, and can only then conclude that they
occur more often on critical days and affected neighbouring days than on any
other day. Yet Falb always refused such an approach.
J.M. Pemter, currently director of the Austrian Meteorological Observation-
Network, has carried out this inquiry and published it in 1892.424 He, for
example, examined the frequency of low pressure systems or barometric
minima in Europe and found out that, during the three years of observations
from 1888-90, each critical day had an average low pressure of 1.67, and on
each of the neighbouring 4 days-according to Falb affected by the critical
days-this figure was also 1.67, as well as on each of the so-called non-critical
days. Therefore, in Europe, low pressure systems are just as frequent on critical
days as on any other day. The same applied to storms whose average occur-
rence was 1.05 on each of the 5 days affected by the critical days, and 1.04 on
any other day. The average number of stations observing rain or snow was 19.4
on a critical day and the same on other days. The total rainfall measured per
day was 132 mm and 138 mm respectively, the number of floods 0.08 and 0.08.
As we can see, the two values are identical everywhere or almost identical, i.e.,
those occurrences that Falb defmed as characteristic for the so-called critical
days are just as frequent on any other day. This outcome is nothing short of
devastating for Falb. It shows that his prognoses have no major value, just like
the one in our example forecasting a sunrise every Thursday. A year does not
have 24 or 25 nor 5 x 24 or 5 x 25 critical days but 365 and in a leap year 366!
Let us now refer in detail to our Swiss weather prophets.
424 Himmel und Erde, IV.
nico.stehr@zu.de
WEATHER PROPHETS 249
Jules Capre in Chillon, the weather prophet of western Switzerland, bases
his weather predictions on the movements of the moon. They are published
annually in the "Almanac des chemins de fer du Jura-Simplon".
Based on the moon's position, Capre predicts the occurrence and duration
of low and high pressure systems for a particular day or a number of days in
certain places of Europe and the resulting future weather conditions for these
regions linking them to cyclones and anti-cyclones-just as modem
meteorologists do. Capre, describing the misfortunes of a weather prophet in
his humorously written preamble to the 1901 prognoses, openly admits that his
1900 predictions were wrong. He blames this failure on the fact that, as an
amateur in the area of astronomy, he was unable to correctly project the
positions of the moon. After having closed this knowledge gap and corrected
the moon's positions he is now confident to be able to achieve better prognoses
in the future. A comparison between prognosis and actual occurrence reveals
again, just as in Falb's prognoses, that they are frequently correct, but often
they are not. Low pressure, rainfall, etc., occur just as often on the predicted
days as they do on days not predicted.
The weather prophet of Aarau, engineer Gladbach, keeps his method in the
dark. Although his Wetterprognosen 42 S, in which he forecasts the weather for a
number of months, contain the "theoretical reasoning and practical training for
observing the barometer with regard to weather predictions", impressing the
amateur with differential equations and research diagrams, Gladbach's method
remains nonetheless unclear. He claims that his diagrams of "Europe's Cloud
Belt", which seem to be the basis of his prognoses yet remain unexplained, are
graphic constructions of the air pressure conditions of previous years. He thinks
it "inopportune" to explain how this is done because of "the danger that an
unauthorised reproduction might be attempted", "discrediting his work".
Gladbach's brief report during a session of the Swiss Society of Natural
Science in Zofmgen in August 1901 again failed to give insight into his
method. Only that much became clear that his work relies on the gravity of the
planets and of the moon.
High school teacher C. Marti, the weather prophet of Nidau, whose prog-
noses frequently appear in German-Swiss papers, also kept his method in the
dark for a long time; he only revealed it in November 1900 during a session of
the Society of Natural Science in Bern, where he handed out printed excerpts,
including a.o. tho an invitation to test his method. 426 Whereas Falb and Capre
base their prognoses on the gravitational and tidal powers, respectively, of the
moon and the sun, Marti refers to a mysterious, completely unknown power.
He assumes that apart from "local constants" and the annual change of the
425 Aarau published by the author.
426 An explanatory account was published after this lecture in the essays of the Society of
Natural Science of Osnabriick.
nico.stehr@zu.de
250 EDUARD BRUCKNER
sun's position weather is affected by the so-called "fast weather factors".
According to his theory, these factors are caused by the planets: If two planets
surrounded by a sufficient layer of atmosphere, one on the near side and one on
the far side of the planetary system, align, that is when rotating around the sun
they come into such a position that a straight line joining the two extends to the
sun, a "disturbance" occurs in that spot on the sun's surface that is facing the
planets. Marti does not disclose the nature of this disturbance. The sun rotates
around its axis, that is once every 26 112 days. The agitated spot rotates with it;
when it faces the earth, it is presumed to cause a disturbance in the earth's
atmosphere. This disturbance is repeated with a further rotation of the sun when
the spot faces the earth for a second time, a third time, in some cases even a
fourth time. Each effective conjunction causes rain or thunderstorms in the
summer, and rainfall or storms in the winter. The most effective conjunctions
are presumed to be those of Mercury and Saturn and Mercury and Uranus,
followed by those of Venus and Jupiter as well as Venus and an asteroid. Marti
has tried to verify his method by examining excerpts from meteorological
yearbooks. Unfortunately his effort suffers from the same methodological error
as Falb's approach: He recounts those cases where in fact storms did occur on
his critical days. This rate of success proves nothing as we have already seen.
In order to justify his method, Marti would have to prove that rainstorms, etc.,
actually occurred more often on those days identified as critical than on any
other day. Marti has so far failed to submit evidence to this effect. Based on an
index of the critical days provided to me by Marti, I have examined the
frequency of rainstorms as defined by Marti, using the weather reports of the
Swiss Meteorological Centre and allowing for the delay in the years 1882-86
and 1894-98 which Marti claimed to have discovered. On any given day of the
entire time span, the average frequency was 0.25, i.e., on one out of four days it
was stormy somewhere in Western Europe. The storm frequency rate on those
days which according to Marti were affected by the conjunction of
Mercury/Saturn was also 0.25, of MercurylUranus again 0.25, of Venus/Jupiter
0.22, of Venus/Juno again 0.25, and of the double conjunction of two pairs of
planets again 0.22. Again the numbers are quite the same and by chance on
critical days even smaller. Storms occur just as often on days identified 'as
critical' by Marti as they'do on non-critical days. Therefore even Marti's
critical days are of no use. This should not be surprising since his entire theory
of a mystical "disturbance" on the sun's surface reflecting onto the earth is
without any physical foundation. If I nevertheless took on the task of checking
his prognoses with great scrutiny, it was done in respect for the energy of a man
who, although in total vain, has invested into his method an enormous amount
of computing.
Therefore, neither Marti's, nor Falb's, Capn!'s, Gladbach's prophecies
stand up to scientific scrutiny. Weather predictions, or rather, weather progno-
nico.stehr@zu.de
WEATHER PROPHETS 251
ses simply cannot be established by ignoring the results of the meteorological
science based entirely on physical evidence and ofthe calculation ofprobabili-
ties.
What does modern meteorology tell us?
The earth's atmosphere as a whole may be compared to a huge machine. Its
heating is provided by solar energy, largely in the tropics; heat loss is highest in
the polar regions. The permanent temperature differences developing between
equatorial and polar regions cause enormous air streams. If we open the door of
a warm room in the winter, we notice-for example with the aid of a burning
candle-that on top, warm air flows out of the room into the open, while at the
bottom the cold air moves into the room. In the same way the warm tropical air
rises and moves towards higher latitudes, while down below cooler air moves
down from higher latitudes towards the equator. This process is affected by the
rotation of the earth around its axis in a manner that we will not elaborate on in
this context.
Streams of air moving side-by-side from different directions or at different
speeds tend to produce cyclonic systems , and frequently to such an extent that
we can recognize the entire convection process in those progressing cyclones.
Any rapid river exhibits this phenomenon: there are descending eddies seen as
a small funnel-shaped indentation in the surface and ascending eddies marked
by gyrating waters; all of these move downstream with the general flow. The
same process takes place in the atmosphere, partiCUlarly in medium and high
latitude regions. As a result of those large prevalent air streams, cyclonic
systems develop at low altitudes, covering an extraordinary large area that quite
often may reach 1500-2000 kilometres and more in diameter. Sometimes these
are ascending cyclones distinguished by low air pressure and are, therefore,
called barometric minima or depressed systems; at other times these are
descending cyclones characterised by high air pressure. The cyclones are
surrounded by winds blowing from different directions that at low pressure, for
instance, consist of the following: west winds at the south side, south winds at
the east side, east winds at the north side, and north winds at the west side. As
the weather is primarily determined by winds, we find the cyclones surrounded
by different weather systems. These cyclonic movements proceed as they are
carried along by the large prevalent air stream in middle and high latitudes
mostly from west to east in successive short intervals of a few days. Their
position is constantly changing in relation to a geographical location and so is
that location's weather.
Careful observation of the skies allows the forecast of imminent changes by
interpreting certain meteorological signs. In our age 'of the telegraph', these
observations include all of Europe. The Meteorological Centre in Zurich, for
instance, is notified during the morning via telegram from different parts of
Europe about the particular weather situation at each weather station on the
nico.stehr@zu.de
252 EDUARD BRUCKNER
same day as of 7.00 or 8.00 a.m. Data about air pressure, temperature, wind
direction, cloud formation, rainfall are reported in this way. Based on these data
that morning's weather map is drawn up for Europe, including those cyclonic
systems previously described. By comparing this weather chart with those of
previous days and relying on previous experience, it is then attempted to
establish what the cyclones' potential position will be the next day. On the
basis of their anticipated course a conclusion is reached for next day's weather
and is published as a prognosis. We are of course today not yet in the position
to determine with mathematical precision where a vortex, which today we
discover somewhere over the ocean near Ireland, will reach us tomorrow. But
within a certain margin it is possible and so are forecasts of the weather. Out of
100 prognoses, about 80 are correct.
Despite the fact that they can only be made for one day in advance, these
prognoses are of practical importance. Proof of this are the gale warnings which
are issued with success at some coastlines as navigational guidelines, and above
all the very precise signal service of the United States, where these forecasts are
particularly important for agriculture.
In Zurich, the prognosis issued by the Swiss Meteorological Centre appears
each afternoon together with the weather chart under the title "Weather
Report. '>427 The public in Zurich also contacts the Centre directly by telephone
and will ask, for example, on a Saturday for Sunday's prognosis. As long as the
personnel are not too occupied, these questions will be answered. It may even
happen that a member of the fairer sex inquires if she should wear a light or a
dark dress for the planned Sunday excursion and the witty officer replies:
"Wear the one that suits you best!"
But among the greater public, the number of followers of weather prophets
the likes of Palb, Marti, and others is still high because the public is hypnotised
by their rate of success. In general, people are not critically inclined. When
education is lacking in this area, observations are done uncritically: successful
predictions stay in mind, failed ones are forgotten. "But he got it right before,"
is the objection. That this "getting it right" is not any different than playing a
lottery with many winnings, is overlooked. Additionally, the public prefers
long-term weather forecasts over prognoses of one day in advance. Weather
prophets cater to this preferencec by announcing their prognoses months in
advance. Even if they are totally useless, they still fmd followers. About this
subject as about many other areas of superstition, it can be said: Mundus vult
427 Besides, the annual subscription of the weather reports through the Swiss mail service is
only 12 SFr.
nico.stehr@zu.de
WEATHER PROPHETS 253
decipi: the world wishes to be deceived. Weather prophets will not die out for a
long time to come.
nico.stehr@zu.de
Chapter 8
An Inquiry About the 35-Year-Period Climatic
Variations·
When in 1889 I finished my research about climatic variations,428 I had to
rely on meteorological observations going no further than 1880 and 1885,
respectively. From the pattern of the rainfall fluctuations I nevertheless
concluded that there had been a maximum of rainfall around 1880 over
global land masses, as well as a minimum around 1860 and a maximum
around 1850. Later in 1895, I described the rainfall variations for Prussia up
to the year 1893 and for the European part of Russia up to 1890 in my article
about the influence of climatic variations on harvests and grain prices in
Europe. 429 In comparison to the years around 1880 rainfall was slightly less
but still above average. After several years have gone by it may now be of
interest to find out whether or not during the past decade rainfall did indeed
decrease all over the globe's land masses. Professor Woeikof has in fact
dealt with this question and counters my findings in his article "D i e
Seespiegelschwankungen zwischen Aralsee und Baraba und die
Brucknersche Hypothese" [The Lake Level Fluctuations between Lake Aral
and the Baraba [steppe in Sibiria] and Bruckner's Hypothesis] published in
the September issue of this journal (1901, p.l99).430 Being occupied with the
final stages of another work, I am unable at present to start a thorough
investigation for all parts of the globe. It is also advisable to at least wait for
the results from the observations of 1900 in order to complete one full
lustrum. Yet, I would like to refer briefly to Professor Woeikow's article
now and, above all, draw attention to a large region where the decrease in
precipitation since 1880 was proven .
• Zur Frage der 35jiihrigen Klimaschwankungen. Petennann's Mittheilungen, 1902.
428 Klimaschwankungen seit 1700 [Climatic Variations since 1700], Wien 1890; see Chapter 4
ofthis collection.
429 Hettner's Geographische Zeitschrift, Volume I, p. 39.
430 Dr. A. Petennanns Mittheilungen, Gotha. Justus Perthes.
255
nico.stehr@zu.de
256 EDUARD BRUCKNER
8.1 WATER LEVEL FLUCTUATIONS IN THE
KIRGHIZ STEPPE AND FLUCTUATIONS OF
RAINFALL IN RUSSIA SINCE 1860.
According to the observations by Berg, Ignatow, and Tangniliew reported by
Woeikof, Lake Aral and other lakes within its vicinity had reached their
lowest level around 1880 and have been rising ever since. A water level
decrease of Lake Aral was also observed over a longer period of time
immediately prior to 1880, yet no details are available as to the beginning of
that decrease. Professor Woeikof states that this pattern contradicts the
climatic variations I claimed to have discovered. In reply, I would like to
point out that, based on the observations available to me in 1889, I had
explicitly declared the Kirghiz steppe, where these lakes are located, an
exceptional area in which the pattern of rainfall variations is reversed, i.e.,
maxima are reached at times when the majority of the other land
masses-3/4 to 4/5 of it-shows minima. 431 I concluded from the situation at
Lake Alakol that "the exceptional region of the Kirghiz Steppe has to be
extended into the corner area between Tarbagatay and Alatau.,,432 If the lake
levels in this region were lowest around 1880 and began to rise thereafter, it
only confirms the exceptional situation of the Kirghiz Steppe, as I stated in
1890.
I do not agree with Professor Woeikof's claim, that the variations of
these lake levels "coincide closely with the rainfall fluctuations at Barnaul. "
What is characteristic at Barnaul is the low precipitation during the years
1859-69, which is entirely consistent with the dry period I had identified for
most of the continental areas. At Barnaul rainfall increases sharply there-
after, while the lake levels, among those Lake Ara1's level in particular, de-
crease and do not reach their lowest levels until around 1880. Consequently,
the minimum lake levels occur 15 years after the minimum rainfall at
Barnaul, and exactly at a time when Barnaul is experiencing a very wet
period. The lakes' fluctuations are much more in accord with the rainfall
fluctuations at the station Irgiz located 200 kilometres further north of Lake
Aral in the centre of the steppe. The following lustra averages (mm) for Irgiz
with the exception of the missing years 1861, 1862, 1884, and 1885, show
this. I have included the related averages for Barnaul.
431 Klimaschwankungen seit 1700 [ClimatiC Variations], p. 170, Table p. 168 f.
432 op. cit., page 176.
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIATIONS 257
1861/65 66170 71175 76/80 81/85 86/90 91/95 96/99
Kirghiz Lake Levels Decrease Lowest Level Increase
Steppe
Irgiz (nun)
Bamaul43
I 193
150
200
173
144
258
148
349
153
338
183
380
197
380 375
Since 1861, precipitation and the lakes' water levels show nearly oppo-
site trends from Barnaul's precipitation. The highest lake levels are mea-
sured at the end of the 90s, a time when Barnaul's precipitation does not rise
further.
Professor Woeikof concludes further that the rainfall at Barnaul does not
correspond with the 35-year climatic variations, and he would suggest a 55-
year period instead. He draws this conclusion from the fact that the centres
of the wet periods at Barnaul supposedly do not coincide with the centres of
the wet periods as I have identified them. I am unable to agree with him in
this. There is no reason to assume that the centre of the first wet period at
Barnaul was reached before the meteorological observations were started, as
Professor Woeikof does. The first observation year of 1838 was dry (251
mm), 1839 was very wet (449 mm), as was 1842 (448 mm). The maximum
occurs at the beginning of the 40s according to available observations; we do
not know what happened prior to that. As to the second wet period, I would
like to point out that the observations at Barnaul are not homogenous
because the rainfall measuring device up to 1882 was placed at 3.1 m above
ground level, thereafter at 1.0 m, and some years even at 2.1 m. This
lowering of the device changed the measurements of precipitation in winter
quite substantially, as can be read from the following comparison:
Time Span Height of the Rainfall Ratio of Summer (May-Sept., 5 mos.) and
Indicator (m) Winter (Oct.-April, 7 mos.) Precipitation
1838-82 3.1 100:49
1883-89 1.0 100: 98
1890-97 2.1 100:83
1898-99 1.0 100: 116
Whether the increase is due to snowdrifts or whether earlier measure-
ments indicated too little snow as a result of the wind's impact remains
unclear. It is obvious though that the measurements of the winter precipita-
tion after 1882 are much too high in comparison to earlier measurements, in
some years even around 100 mm and more.
By smoothing these values434 and by determining a 62-year mean (289.8
mm), we arrived at the following periods whereby each 5-year average refers
433 See further down about the adjustment to the winter precipitation that was done because of
the lowering of the rainfall measuring device at Bamaul after 1882.
nico.stehr@zu.de
258 EDUARD BRUCKNER
to the centre year of that time span. For comparative purposes I have
included the climatic variations of wet and dry time span which I had
established in 1890. 435
Precioitation at Bamaul Climatic Variations as Established in 1890
Above the Mean 1840--49 Wet 1841-55
Below the Mean 1850-73 Dry 1856-70
Above the Mean 1874-97 Wet 1871-85
There is a perfect compliance with the wet and dry periods of the climatic
variations as far as those have been traced for the majority of countries, i.e.,
up to 1885. However, at Bamaul even after 1885, precipitation remains high.
Whether this is a single event or is repeated in other regions is now the
question. According to the following tabulation (Table 8.1) the latter seems
in fact to be the case for Russia, at least in part.
Table 8.1. Deviations of the ten-year averages of precipitation (mm) from the mean in the
river re~ions of the EuroEean Eart of Russia
Upper Central Lower Upper Lower Don Mean
Wolga Wolga Wolga Dnjepr Dnjepr of River
andOka and Kama Regions
Mean 513 462 354 533 435 450 458
1861-70 1 -28 -14 5 -32 -15 -14
1862-71 -2 -33 -23 8 -37 -13 -17
1863-72 7 -32 -9 18 -28 -20 -11
1864-73 19 -12 10 25 -19 -4 3
1865-74 23 1 21 11 -18 -4 6
1866-75 19 16 36 5 -19 2 10
1867-76 27 22 37 15 -4 11 18
1868-77 16 16 35 10 5 23 18
1869-78 29 16 31 18 15 27 23
1870-79 26 23 37 22 24 42 29
1871-80 29 29 38 6 18 40 27
1872-81 25 25 48 -6 15 51 26
1873-82 11 19 41 -5 18 58 24
1874-83 7 3 30 3 19 54 19
1875-84 1 -2 30 12 20 58 20
1876-85 4 -11 18 7 19 46 14
1877-86 -1 --14 20 -3 23 44 12
1878-87 5 -3 14 -1 14 33 10
1879-88 11 -2 14 -10 13 24 8
1880-89 8 1 11 -13 4 10 4
1881-90 -11 3 1 -9 -1 -1 -3
1882-91 -14 3 -8 -2 -3 -22 -8
[continued on next page]
434 Adjustment factor of the winter precipitation 1883-89 and 1898-99,49:98 = 0.5; 1890-97,
49:83 = 0.6. Without this adjustment the average would be 317 mm, and the years
1840--42 as well as 1876-97 would show above average values.
435 See also in Klimaschwankungen [Climatic Variations], p.192.
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIATIONS 259
[continued f!:.om e.revious e.agel
Upper Central Lower Upper Lower Don Mean
Wolga Wolga Wolga Dnjepr Dnjepr of River
andOka and Kama Regions
Mean 513 462 354 533 435 450 458
1883-92 -12 7 -14 -4 0 -31 -9
1884-93 -6 18 -10 2 6 -28 -3
1885-94 6 25 -19 13 11 -30 1
1886-95 8 24 -17 23 19 -22 6
1887-96 14 23 -16 32 19 -21 8
1888-97 -4 5 -24 26 15 -30 -2
1889-98 -9 -4 -30 25 11 -32 -7
The table shows the deviations of the averages of successive ten-year
periods436 in relation to the mean of all years and is based on the figures for
the total annual precipitation provided by E. Heintze in his essay about the
deviation of precipitation from the mean in the river regions of European
Russia during the time span of 1861-98. 437
The raw rows of data sufficed to convince E. Heintze that his findings
about the fluctuations of precipitation, in general, come pretty close to my
own results. The smoothed rows show this even more clearly: With one
exception (region of the upper Dnjepr) all river regions show a pronounced
maximum of precipitation in the decades of 1870-79,1871-80, 1872-81, or
1873-82, which was preceded by a minimum in the 60s. Precipitation begins
to decrease considerably in the 80s, showing a sharp and well-pronounced
decline at the lower Wolga and the Don respectively; it is less noticeable at
the central Wolga and lower Dnjepr, where precipitation remains fairly high
up to 1889/98 even though it is lower than in the 70s. Thus only the central
sections of the Wolga river and the Dnjepr region show a similar pattern of
rainfall as at Barnaul. In contrast, the other parts of Russia show a consider-
able decrease of rainfall after 1885, as well as East Siberia's Nertschinsk,
and Nikolajewsk at the Amur. The smoothed figures for the rainfall observa-
tions in these latter places are included in Table 8.5 below, as well as those
of most of Central Europe represented here by the two cities of Bruxelles
and Bremen.
436 The multi-year averages were re-determined for each single year according to the
instructions by Heintze.
437 Meteoro1ogische Zeitschrift, 1901, p. 219.
nico.stehr@zu.de
260 EDUARD BRUCKNER
8.2 DECREASING RAINFALL IN THE UNITED
STATES SINCE THE MIDDLE OF THE 80S.
One area that is known to suffer a general decline in precipitation since 1888
is that of the United States.
Alfred J. Henry, departmental director with the U.S. Weather Bureau,
discovered while examining rainfall patterns in the U.S.,438 that it was too
low from 1887-96 in almost all of the country. Based on his data, I put
together the following table, which includes the deviation of precipitation,
that is the deviation from the multi-year means in each of the different parts
of the United States for the decade of 1887-96. Henry provides the
deviations for each year, which limited space does not allow here. To show
the relative extent of the deviations, they are also expressed in percentages of
the multi-year averages (1870-88) which Hann439 provided for the individual
regions. These percentages are, of course, estimates because Hann's data
refer, at least in part, to somewhat different stations than those used by
Henry. Nonetheless these percentages highlight the lack of rain in the arid
regions.
Only the New England states and the central Atlantic states, the southern
part of the eastern slope' of the Rocky Mountains, the Pacific coast, and
California show extensive or normal rainfall. The rest of the regions show
too little rainfall. This lack of rain is particularly evident in the Gulf states
and the upper Mississippi valley; in the Gulf states, for instance, rain
decreased by 105 mm and 104 mm respectively or 7 and 9 percent compared
to the multi-year average. A large decrease is also evident in the percentages
from the upper Great Lake region, the Missouri valley, the central part of the
Rocky Mountains' eastern slope as well as on their southern plateau. Henry
emphasises correctly that, as far as the United States are concerned, too little
rain in one region is not compensated by too much rain in another. In fact,
Table 9.3 shows that in only one year (1896) out of the ten were there more
regions with too much rain than with too little, while in four years three
quarters of the regions suffered too little rain all at the same time.
438 Report of the U.S. Weather Bureau for 1896/97, p. 328, Washington 1897.
439 Compare Klimatologie [Climatology], 2. Edition, Vol. III, p. 289
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIATIONS 261
Table 8.2. Deviations of rainfall 1887-96 in the different re~ions of the United States
Regions (Districts) Number of Deviations Number of Years with:
Stations from Normal
(mm) (%) Too Little Too Much
Rainfall Rainfall
New England 9 +1 0 5 5
Central Atlantic States440 II +12 +1 5 5
Southern Atlantic States II --66 -5 7 3
Eastern Gulf States 5 -105 -7 9
Western Gulf States 8 -104 -9 8 2
Ohio Valley and Tennessee II --68 --6 7 3
Lower [Great] Lakes Region 8 -38 -4 7 3
Upper [Great] Lakes Region 10 --68 -8 9 1
Outer North West 4 -18 -4 8 2
Upper Mississippi Valley 13 -83 -9 7 3
Missouri Valley 10 -52 -7 7 3
Mountain Range-East/North 6 -2 -I 5 4
Mountain Range-East/Central 7 -43 -8 7 3
Mountain Range-East/South 4 +6 +1 4 6
Mountain Range-Northern 4 -21 -5 6 3
Plateau
Mountain Range--Central Plateau 5 -5 -2 5 5
Mountain Range--Southern 9 -23 -8 5 5
Plateau
Northern Pacific--Coast 8 +41 +4 5 5
California 12 +21 +2 4 6
Table 8.3. Number of Regions with
Too Little Too Much Too Little Too Much
Rainfall Rainfall Rainfall Rainfall
1887 15 4 1892 II 7
1888 10 8 1893 12 7
1889 14 5 1894 15 4
1890 10 9 1895 15 4
1891 10 9 1896 8 11
The above figures are deviations from the multi-year mean; the ten years
from 1887-96 were too dry compared to the standard value. More extreme is
the difference to the immediately preceding decade, which on the average
was too wet. This is shown in the next table which, based on Henry's data,
includes the ten-year averages of 1877-86 and the deviations of the averages
of 1887-96 from those means by the [U.S.] Weather Bureau. I also added
the mean values of 1897-99 determined on the basis of the Weather
Bureau's report.
440 This category is almost identical to the category "Southern Atlantic States" in my book
because I did not include stations located south of Washington [D.C.].
nico.stehr@zu.de
262 EDUARD BRUCKNER
During the decade of 1887-96, all stations, with the sole exception of
New Bedford, Mass. and Augusta, Ga., had less rain than during the
previous decade, and in some places even substantially less. Towards the end
of the century the aridity has increased even more, so that in 1897/99 the
deviations are considerably higher still. As the deviations of 1877-86 are
higher than the above-mentioned standard deviations, it must be concluded
that the decade of 1877-86 was too wet.
The tabulated figures clearly show a maximum rainfall in 1877-86
followed by a considerable decrease that has been observed until 1899. Only
the New England states and the neighbouring central Atlantic states do not
fit this pattern. Instead these states show an increase in rainfall and must
therefore be considered exceptional regions among the rest of the U.S., as I
had discovered earlier for the time around 1880.
Table 8.4.
Station Mean Deviations from the Mean 1877-86
1877-86 {mm~ {%~
{mm~ 1887-96 1897-99 1887-96 1897-99
New Bedford, Mass. 1133 +127 +206 +10 +18
Providence, Rh. I. 1290 -17 +68 -1 +5
Lenoir, N.C. 1305 -66 +31 -5 +2
Hatteras, N.C. 1865 -398 -435 -21 -24
Wilmington, N.C. 1494 -293 -467 -20 -31
Charleston, S.C. 1443 -113 -235 -16
Augusta, Ga.. 1182 +23 +42 +2 +4
Savannah, Ga. 1293 -55 +31 -4 +2
Jacksonville, Fla. 1465 -217 -237 -15 -16
Mobile, Ala. 1692 -200 -208 -12 -12
Montgomery, Ala. 1322 -55 -157 -4 -12
Vicksburg, Miss. 1576 -390 -313 -25 -20
Memphis, Tenn. 1442 -203 -311 -14 -22
New Orleans, La. 1558 -179 -512 -11 -33
Shreveport, La. 1378 -351 -518 -26 -38
Galveston, Tex. 1331 -316 -374 -24 -28
8.3 RAINFALL FLUCTUATIONS FROM 1830 TO 1900
IN THE UNITED STATES, AS WELL AS AT SOME
STATIONS IN CENTRAL EUROPE AND EAST
SIBERIA.
Henry, who appears to be unaware of my own research about climate varia-
tions and, hence, seems to be unbiased in his observations, raises the
question of the rainfall pattern in former times. In an effort to demonstrate
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIAnONS 263
this, he includes in his table the observations from three stations in New
England (Boston, New Bedford, and Providence), three in the upper Ohio
valley (Marietta, Portsmouth, and Cincinnati) and four in the central
Mississippi valley (Muscatine, Monticello, Marengo, and Peoria). Henry
smoothed these numbers according to the following formula:
(a +4b + 6c +4d+ e) 116 =c'
For each of the three geographical areas, he then determines a mean
value. These figures, which he also expresses in graphical curves, clearly re-
flect climatic variations in the Ohio and Mississippi valleys, while the New
England values deviate.
Still, there remain certain secondary distortions because the above
formula takes only three years into account and factors "a" as well as "e"
have very little weight. I have made further adjustments by combining five
years each of Henry's smoothed figures into one mean; regrettably I was
unable to go back to the original observational data for each of these years
because Henry does not provide them. Fully adjusted, the formula reads:
(a+5b+ llc+ 15d+ 16e+ 15f+ llg+5h+i)/80=e'
In accordance with his report, I have supplemented the data sequences up
to the year 1899, enabling a smoothing until the year 1897.
I am adding the data sequences for Bruxelles and Bremen which have
been smoothed by forming 5-year averages, i.e., for 1832 the lustrum
1830-34, for 1833 the lustrum 1831-35, etc., and which are the only stations
in Central Europe for which observations of the year 1900 are presentlYWI
available to me, as well as the figures for the ore smelter of Nertschinsk and
Nikolajewsk at the Amur in Siberia. The adjustment was done on figures
rounded to full centimetres. Missing monthly measurements by the stations
Nertschinsk and Nikolajewsk were interpolated beforehand according to
those months' precipitation in the four neighbouring years. If an entire year
was missing, four- instead of five-year averages were used. All numbers are
in millimetres.
Aside from shorter oscillations that Henry concentrates on, there are
sharply defined rainfall variations evident in the upper Ohio and the central
Mississippi valley. Aridity prevails in the middle of the 1830s; rainfall then
increases and reaches a maximum toward the end of the 1840s, decreases
again and stays about average for some time arriving at a minimum at the
beginning of the 1870s; a rapid increase to another maximum in the early
1880s is then followed by a considerable decrease towards the end of the
century. This pattern is in complete accordance with the periods of climatic
variations as I have established them earlier. Only the minimum of 1871172
occurs quite late at a time when, in most areas, rainfall is already slightly
441 December 1901.
nico.stehr@zu.de
264 EDUARD BRUCKNER
Table 8.5.
New Upper Central Bm- Bre- Nertschin- Nikola:
England Ohio Missis- xelles men sky Smelter jewskat
(Non Appl. Valley sippi Plant theAmur
Region} Valle~ {Siberia} {Siberia}
Mean 1163 1049 894 725 690 413 454
Value
1830 38
1831 20
1832 -23 5 -24
1833 -84 -41 -{i8*
1834 -145 -89 -45 18
1835 -190 -99 -33 62
1836 -208* 107* -75* 90
1837 -201 -84 -9 88
1838 -175 -76 -5 74
1839 -142 -48* -15 58
1840 -114 -36 -37 28
1841 -99 -8 3 52 -17
1842 -102 46 7 70 41
1843 -122 41 39 108 77
1844 -147 66 9 54 109
1845 -160 97 5 52 159
1846 -163 124 5 56 122
1847 -152 155 -17 56 75
1848 -119 168 -9 64 35
1849 -81 145 17 108 13
1850 -53 99 75 130 -15
1851 -28 61 47 68 -35
1852 -18 25 55 60 -23
1853 -28 -5 272 21 52 -53
1854 -43 -23 150 27 24 -9
1855 -41 -18 66 -{il -48 -17
1856 -30 8 48 -91* -80 -33
1857 -15 30 64 -87 -80 -57
1858 0 41 74 -59 -50 -{i3
1859 23 51 76 -{i3 -38 -103
1860 46 43 66 -21 34 -117·
1861 58 8 41 -3 88 -101
1862 58 -15 18 -53 18 -103 -124*
1863 53 -5 10 -91* -82 -87 -116
1864 51 18 -5 -87 -104 -{i7 -116
1865 53 30 -25 -59 -88 -51 -82
1866 71 48 -23 -35 -92 -41 -72
1867 99 43 -5 39 -42 -43 -102
1868 122 5 -10 41 -20 13 -106
1869 130 -53 -23 9 -48 -7 -118
1870 124 -91 -33 7 -90 26 -102
1871 104 -109· -53 13 -136 7
1872 86 -107 -71* -25 -154* 13 -112
1873 81 -81 -51 -25 -136 -5
1874 84 -46 -8 11 -104 21 46
1875 91 -30 33 19 -74 -27 47
1876 99 -33 5.8 101 -16 -17 101
[continued on next page]
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIATIONS 265
[Continued f!.om e.revious e.ogel
New Upper Central Bru- Bre- Nertschin- Nikola-
England Ohio Missis- xelles men sky Smelter jewskat
(Non Appl. Valley sippi Plant theArnur
Re&ionl Valle~ ~Siberial ~Siberial
Mean 1163 1049 894 725 690 413 454
Value
1877 94 -30 78 115 36 15 94
1878 76 -15 91 155 88 31 130
1879 61 18 97 147 98 65 36
1880 36 74 97 139 120 119 38
1881 10 119 102 67 112 135 26
1882 8* 127 112 69 136 115 14
1883 18 97 94 37 94 59 11
1884 33 51 53 31 82 7 6
1885 38 28 10 -33 112 -27 28
1886 104 -28 -25 -1 -6 -55 28
1887 142 -18 -56 9 -50 -55 47
1888 168 15 -71 35 -66 -25 50
1889 170 38 -69 19 -40 -43 56
1890 142 38 -56 19 -24 -37 18
1891 97 13 -53 -33 -24 -7 2
1892 46 -41 -66 -33 -6 -19 -10
1893 0 -99 -81 -53 8 -3 -18
1894 -18 -130* -102 -45 -30 33 -6
1895 -7 -109 -107 -23 34 23 -16
1896 18 -76 -109* -33 -16 7 -30
1897 36 -30 -104 -49 -58 -23 -80
1898 -67* -74*
above average. Nonetheless, the maximum of 1882 complies again with my
findings.
The periodicity of a complete variation reaching from the minimum of
1836 to the minimum of 1871 is 35 years, and from the maximum of 1848 to
the maximum of 1882, 34 years.
The New England states deviate completely from this pattern showing an
arid period from the early 1830s to the early 1850s, a maximum around 1869
and a secondary minimum around 1882. The New England states are
exceptional regions, as explained earlier in my book. But the most striking
feature of the New England data is the increase in rainfall that manifests
itself as early as 1816 and interferes with the observed variations. This
increase is evident in all of the available long-term data rows. It has never
been fully explained. Could it be symptomatic of a long-term climate period,
such as the 160-year periodicity that Sieger442 has shown to be probable for
Scandinavia?
Bremen and Bruxelles, representing the Northwest corner of Central
Europe, show similar fluctuations: a minimum in 1833 and 1836
442 Zeitschrift der Gesellschaft fUr Erdkunde, Berlin 1893, p. 444.
nico.stehr@zu.de
266 EDUARD BRUCKNER
respectively, a maximum in 1850, in Bremen a minimum in 1872, and in
Bruxelles in 1856-63, a maximum in Bremen in 1882, and in Bruxelles
1882. Bremen shows the same delay in the occurrence of the minimum of
the 60s as was observed in the central parts of North America, while in
Bruxelles the maximum (1878) occurs slightly earlier. In general, the trend
of the curves is the same.
That is also the case for Nertschinsk in East Siberia and Nikolajewsk at
the Amur. At Nertschinsk the first maximum (1845) occurs some years
earlier than in the interior of the United States, at Nikolajewsk as in the
European part of Russia it is the second maximum (1878).443
The interior of the United States, Bremen, Bruxelles, the European part
of Russia, Nertschinsk, and Nikolajewsk at the Amur, all these locations
show an obvious 35-year periodicity of climatic variations, while there is no
trace of a 55-year period. The epo,chs vary in some cases slightly-they
occur earlier or later. The early or late occurrences will however be evened
out with the next epoch. The time frames of the wet and the dry periods
show similar irregularities. I am adding a table about the dry and wet periods
at this point.
d~ wet d~ wet dry
Ohio-Valley 1833-41 1842-52 1853-78 1879-91 1892-97
Mississippi-valley 18? --63 1864-74 1875-85 1886-97
Bruxelles 18? -40 1841-54 1855--66 1867-90 1891-98
Bremen 18? -33 1834-54 1855-76 1877-85 1886-98
Europ.Russia 18? --67 1868-84 1885-?
Nertschinsk 1842-49 1850--67 1868-84 1885-97
Nikolajewsk a. A. 18? -73 1874-91 1892-97
Centro Germ. border 18? -38 1840-54 1855-71 1872-87 1888-?
Bruckner 1890. 1826-40 1841-55 1856-70 1871-85
If we look at the average years in which the periods ended we almost get
the same results, i.e., with deviations of not more than two years, as I had
established in 1890.
The discrepancies in the individual regions should not surprise us: we are
dealing with meteorological periods and not with mathematical ones. Just as
the daily temperatures vary from one day to the next and differ in their
extremes and length of time, just as their maximum and minimum shifts
according to cloud cover, wind, etc., in short according to weather condi-
tions, so do climatic variations. Only the use of a large amount of data, as I
have done in my book, can help to eliminate the local and temporary effects
443 Op. cit.
nico.stehr@zu.de
35-YEAR PERIODS OF CLIMATE VARIATIONS 267
of random occurrences. 444 1t should be strictly differentiated between these
temporary irregularities of unknown causes and the lasting exceptional dis-
crepancies; these refer to regions where the pattern of the oscillations is
more or less reversed. An inquiry into these exceptional regions in particular
might shed some light on the mechanism of climatic variations.
Let us take a look at the amplitude of these variations expressed in
percentages of the multi-year averages for the different rows of data and
compare it with the figures which I arrived at earlier""5 using a different
method (through lustra averages) and dealing largely with a totally different
subject matter.
Average Average Variations My Results of
Minimum Maximum 1890
Lower Ohio area -11% +14% 25% } 19%
Central Mississippi-region -10 +13 23
Bruxelles -II +16 27 } 20%
Bremen -16 +20 36
Nertschinsk -28 +36 64 } 47%
Nikolajewsk -28 +29 57
According to the new figures, the oscillation is considerably larger, espe-
cially in those data rows provided by one station only. This cannot be
otherwise: a row of data that, for each single year, includes a five-year mean
must show a larger oscillation than the same row based on lustra averages
alone. These variations of the rainfall are in fact slightly larger than I had
identified them in 1890 with the help of lustra averages.
Let me summarise briefly what I described above:
1. Professor Woeikofs findings about the variations of Lake Aral and its
surrounding lakes confirm my conclusion of 1890 that the Kirghiz Steppe
is an exceptional case with regard to climatic variations.
2. The rainfall fluctuations at Barnaul do not correspond with the
fluctuations at these lakes but show instead a 35-year periodicity of
alternating wet and dry climate periods. Noticeable is the long duration of
the wet period at the end of the century.
3. In contrast, East Siberia and the selected Central European stations in
most parts of European Russia show a considerable decline in rainfall
towards the end of the century.
4. Significant is the decline in precipitation in the United States since the
middle of the 80s. Regions exempt from this phenomenon are
444 Please compare Klimaschwankungen [Climatic Variations], pp. 175 and 192, regarding
these irregularities.
445 We are unable to supply data for the European part of Russia because the absolute mini-
mum of 1860 is not included in our table.
nico.stehr@zu.de
268 EDUARD BRUCKNER
predominantly the New England states, as well as the central Atlantic
states, which is in full accordance with my earlier assumptions.
5. The 35-year periodicity of precipitation is also clearly evident when the
data are smoothed as suggested by P. Schreiber446 by determining
progressive group averages. The resulting figures are identical with my
own findings which I gained earlier through the shorter method of lustra
averages; except that the fluctuation of the rainfall is slightly larger than
in my study.
6. Occasionally the timing of the epochs is irregular; climatic variations are
after all a meteorological phenomenon and not a mathematical one.
7. None of the observed rows show any trace of a 55-year fluctuation.
446 Excerpts of the Koniglich Sachsisches Meteorologisches Institut Leipzig 1896, Issue 1,
p. 46. (Also in Civilingenieur, Volume 42, Issue 1 and 3.)
nico.stehr@zu.de
Chapter 9
About Climate Variability*
The common interest in the question of changes in the climate of smaller or
larger regions can be explained by the fact that all life on earth, especially
the plant life of a region, depends on the climate to a large degree.
Temperature and rain primarily have a decisive impact on plant life, each
playing a slightly different role. Because each plant requires a certain
amount of warm temperature in order to exist, which may differ from species
to species, and since the summer temperatures decrease closer to the polar
regions, it is quite obvious that temperature is a particularly significant factor
in determining the polar line of plant growth. For example, an imaginary line
connecting all those spots where the temperature rises to more than 10°
during three months in the summer roughly coincides with the polar tree
line. The amount of precipitation, on the other hand, often determines plant
life in its growth from West to East. Due to the fact that in the Old World
rainfall decreases from the Atlantic Ocean in an easterly direction, plants
requiring a lot of moisture are limited to western regions and those which
thrive only when there is little moisture, to the East.
If weather conditions would remain the same year after year, the growth
line of a plant species would be indicated world-wide, and all other
conditions being equal, by the same isotherm, i.e., the line connecting places
with the same average temperature, or the same isothyte, i.e. with the same
annual amount of precipitation. Of course this is not the case, because any
extremes of the two factors are much more important for the distribution of
plants than average temperatures and average precipitation recorded over
many years of observations. Over several years, a forest-tree in a certain
location may have received a sufficient amount of moisture to grow on, yet
if in the following year the required minimum amount is not reached, the
tree and its seedlings will perish. This is a common fact. In a series of good
years a plant will be able to extend its growth area; yet the first bad year
• Uber Klimaschwankungen, Mitteilungen def Deutschen Landwirtschaftsgesellschaft, 1909.
269
nico.stehr@zu.de
270 EDUARD BRUCKNER
which fails to supply the warm temperatures and the moisture needed for its
survival, will force it back to its original habitat. In the long run, plants are
only able to survive in areas where even in bad years the required minimum
temperature and moisture is provided. Obviously the weather pattern of bad
years determines the boundaries of the plant's growth area. This primarily
refers to the natural habitat of the plant.
Wherever humans interfere, conditions are partially altered. Often it can
be an advantage to grow a plant that thrives only in good years and fails to
produce or perishes in bad years. The yield of one good year may
compensate for many bad years. For example, in the Middle Ages and later,
wine was grown in many regions of Northern Germany and Northern
France, where nowadays viniculture is out of the question, even though some
years did not produce at all. Because of high freight costs, it was more
advantageous to put up with poor crops than to import grapes from the south.
Later on as infrastructure improved and transportation became cheaper, and
also the taste became more refined, wine-growing moved southward and
vineyards in the North were re-cultivated with a product more suitable for
that climate.
A similar development is taking place today in the area of grain
production.
Temperature conditions determine how far north the different types of
grain are grown. Temperature restricts wheat growing in Europe primarily to
regions south of the 60th or 61 st parallel of latitude, while allowing the
growth of the unexacting grain types, such as barley and oats, almost as far
north as the Northern Cape and the White Sea. Nonetheless, in the main
agricultural regions of Europe temperature has only a limited local impact
and its influence on the outcome of the crop is therefore secondary. Only
close to the polar boundary of the grain's growth area, e.g., in Scandinavia
and Northern Russia, crops are often damaged by frost. Too much heat does
not damage grain as long as enough water is available, e.g., wheat grows
well in the wet tropical part of India's subcontinent!
Moisture is a much more important factor in growing wheat. Without
water, there would be no agriculture, and with too much water, none either.
This rule comes to mind when we look at the expansion of agriculture
around the globe or rather when we examine the causes of crop failures.
Droughts and poor harvests go hand in hand in all those regions with little
rainfall. Where rainfall is plenty, poor harvests are mainly evoked by years
of heavy rainfalls. Of course, it is not the amount of rainwater per se that
determines this, but the ratio between rainwater and evaporated water. The
same quantity of water that inevitably drowns the crop in a cold climate may
just be sufficient enough for the irrigation of grain fields in a hot climate. In
contrast, a low amount of rainfall causing droughts in a hot country due to
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 271
heavy evaporation might often be entirely sufficient in a cool climate. For
Europe's wet coastlines of the North Atlantic Ocean and Central Europe, the
effect is the exact opposite to the dry interior of the continent. Southern
Russia on the one hand and Great Britain and Ireland on the other are
extremes in this regard. In the majority of cases, droughts are the causes for
crop failures in southern Russia, whereas the many crop failures that
England endured at the end of the 1830s, during the 1840s, and at the
beginning of the 1850s, as well as the 1870s and 1880s, occurred in years
with excessive rainfalls.
Southern Europe and most of the Tropics, at least as far as they are
producing grain, experience similar conditions as in Southern Russia. The
years of famine in the East Indies coincide with dry years.
Central Europe, and the German Empire in particular, is halfway between
these two extremes. Dry years are often good years for a number of agricul-
tural activities, especially wine and fruit growing, but also for grain,
specifically in the West, whereas they are bad years for dairy farming or
cattle raising. Although occasionally dry weather does damage crops, crop
failures due to too much precipitation are much more common. A detailed
inquiry into the quite complicated causes of crop failures is still outstanding.
Under the climate factors' determining influence on crop failures, a
radical shift in agricultural production has taken place over the last decades
starting around the middle of the past century. Regions where crop failure
used to be a frequent occurrence because of too much rain have mainly
abolished grain cultivation and have almost exclusively turned to the cultiva-
tion of grasslands and of fodder-plants instead, both of which are perfect for
a wet climate and a prerequisite for keeping and breeding cattle. This can be
said of Ireland and England, but also of Denmark and of larger regions of the
German Empire, in particular the northwestern regions (e.g., Holstein,
Friesland, and Western Prussia). Due to the modernised transportation
system in these regions, the supply of grain from other regions became easier
and allowed agricultural production to adjust much better to the climatic
conditions. This is also the case for Switzerland.
In view of the close link between plant life and weather and climate, their
influence on the outcome of harvests is widely undisputed. If for that matter
the climate was constant and the weather changed from year to year
according to random law, then, on the average of a long series of years, the
harvested crops would, on the whole, have to be constant as well. Is the
climate actually constant?
Recent studies have shown that this is not the case; in fact climate has
many fluctuations that vary around a middle range. I am not referring to the
II-year weather cycle that has often been claimed yet never been verified
completely, which disappears entirely from time to time and is caused by the
nico.stehr@zu.de
272 EDUARD BRUCKNER
ll-year cycle of sunspot activity, but rather to the far more significant 35-
year climate variations, the probability of which I was able to declare in
1888 and which I could verify for larger parts of the globe in 1890.447 Since
then I have continued with my research and summarise as follows.
The climate variations involve multi-year changes in temperature, air
pressure, and rainfall that take place simultaneously all over the globe. The
temperature is the one factor in this process upon which all others depend:
Temperature variations are common to almost all regions on earth. As shown
in Table 9.1, they all experience cold and warm periods at the same time.
Figures for the three continents Europe, Asia, and North America indicate
the deviations by 11100 degree Celsius of the five-year (lustrum) averages
from the multi-year (30-50) average. The minus symbol (-) stands for a
lustrum that was too cold by the indicated degree and the positive figures (+)
for lustra which were too warm.
Table 9.1. Temperature Variations of the Three Continents of the Northern Hemisphere
1806-1885. (11100° C)
Five Year North
Lustrum EuroEe Asia America Globalll
1806110 -05 -16 -18
1811115 -34* -38* -46
1816/20 -05 -37 -35
1821125 +46 +54 +29 +56
1826/30 -03 +17 +93 +14
1831135 +14 -01 -18 +03
1836/40 -54'" -31* --63'" -39
1841145 -08 +09 ~17 +00
1846/50 +04 +02 +17 -08
1851155 +03 +29 +16 +11
1856/60 -08 -05 -28 +06
1861165 +11 -07 +29 -06
1866/70 +26 +31 -07 +11
1871/75 -02 +25 +04
1876/80 -09 -05 -07
1881185 -08 -08
It is evident from these figures that Europe, Asia and North America
experienced a relatively cold period from 1806 to 1820, followed from 1821
447 Compare Ed. Bruckner, Klimaschwankungen seit 1700 [Climatic variations since 1700;
Chapter 4 o/this anthology] Wien 1890. - Der EinflufJ der Klimaschwankungen au/ die
Ernteertriige und Getreidepreise in Europa [Influence 0/ climate variability on harvest
and grain prices in Europe; Chapter 6] Geographische Zeitschrift, I, 1895.- Zur Frage
der 35jiihrigen Klimaschwankungen [An inquiry about the 35-year period 0/ climate
variations; Chapter 8], Petermann's Mittheilungen 1902, issue 7.-Schwankungen des
Niederschlags im Deutschen Reiche [Variations o/precipitation in the German Empire],
Zeitschrift fUr Gletscherkunde, fUr Eiszeitforschung und Geschichte des Klimas. 1 (1906).
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 273
to 1835 by a fairly warm one, and then alternating between cool (1836-45),
warm (1846-75), and cool (1876-85) periods. In all continents the maxima
of the cold periods, marked by an asterisk (*), fall consistently into the
periods 1811-15, 1836-40, and (to a lesser extent) 1876-85. The maxima of
the warm periods, highlighted in bold, occur with the same consistency in
Europe and Asia in 1821-25 and 1866-70, while in North America they shift
by five years (1826-30 and 1861-65). Space does not permit to include the
data for the Southern Hemisphere; they however show a similar picture. I
have not yet followed up on global temperature variations after 1885.
Temperatures vary on the average by almost 10 C. That is quite substan-
tial. For it means nothing else than that during the five years of 1836-40 the
average annual temperature of Berlin, for instance, was a full degree lower
than during the five years of 1821-25, as if that city from one period to the
other had moved further north by 3 degrees latitude. During the last 30
years-since approximately 1875-the temperature fluctuations have
become less distinct and, therefore, appear less conspicuous at individual
stations.
The distribution of air pressure is affected by temperature variations. In
Europe, as we will later find out in detail, the transfer of moist air from the
ocean to the continent seems more difficult during warm periods, yet easier
during cold periods. This in turn must affect the rainfall of the region. And
indeed, rainfall fluctuations are much more pronounced and can be traced
much better than temperature fluctuations. Mind you, the rainfall measure-
ments of one single station fail at first to show a pattern of regularity from
one year to the next; rain is, after all, an extremely unsettled meteorological
phenomenon. One or two major thundershowers coming down in one parti-
cular spot can significantly affect that location's annual amount of precipita-
tion. It is advisable to combine the measurements of a number of stations for
a somewhat larger area. This should be done in such a manner, that for each
station the difference between the amount of precipitation during the
observed year and the multi-year average of rainfall is expressed in
percentages indicating whether too much (+) or too little (-) precipitation
occurred. In addition the average deviation is then determined for all the
stations combined. In the table below this has been done for the large river
regions of the German Empire.
Table 9.2. Variation of Precipitation for the River Regions of the German Empire: Deviation
of the Amount of Precipitation in Percentage ofthe SO-year Mean Value
Year Weichsel-Oder Elbe Lower Rhein Upper Rhein Mean
I~I ~ ~ ~ ~ ~
1852 +3 +17 +24 +5 +12
1853 +5 ±O -2 -3 to
1854 +17 +9 +13 +5 +11
1855 +12 -3 -5 -2 ±o
nico.stehr@zu.de
274 EDUARD BRUCKNER
Year Weichsel-Oder Elbe Lower Rhein UEEerRhein Mean
1856 -3 -8 +10 +12 +3
1857 -31 -31 -32 -34 -32
1858 -19 -13 -20 -21 -18
1859 -12 -5 +13 -3 -2
1860 -8 +15 +17 +20 +11
1861 +3 +4 -3 -7 -1
1862 -15 +8 +3 +7 +1
1863 -13 -2 -8 -11 -9
1864 -8 -16 -25 -28 -19
1865 -12 -25 -12 -27 -19
1866 -2 +9 +21 +16 +11
1867 +23 +14 +17 +16 +18
1868 -9 +4 -4 +4 -1
1869 +4 +2 iO -3 +1
1870 -7 +1 -2 +2 -2
1871 -9 -5 -4 -13 -8
1872 +9 -4 +13 +3 +5
1873 -II -15 -15 -8 -12
1874 -21 -21 -15 -12 -17
1875 -I -2 +15 +11 +6
1876 -2 +8 -1 +5 +3
1877 +7 +14 +25 +15 +15
1878 +1 +2 +14 +26 +11
1879 +9 +14 +7 +8 +10
1880 +14 +22 +17 +20 +18
1881 -10 -I -4 -10 --6
1882 +18 +26 +34 +36 +28
1883 +5 -8 -5 --6 -3
1884 iO +4 -12 -12 -5
1885 +8 --6 -I +7 +2
1886 -13 -7 -I +4 -4
1887 --6 -17 -14 +17 -13
1888 +16 +8 +13 +4 +9
1889 +9 +5 -5 iO +2
1890 +10 -3 -2 iO +1
1891 +14 +6 iO -I +5
1892 -13 -18 -14 -7 -13
1893 +4 --6 -11 -13 --6
1894 +5 +13 iO -3 +4
1895 +1 -2 --6 +2 -1
1896 -3 -1 -4 +10 iO
1897 +5 +3 -8 -4 -I
1898 +8 +2 -10 -4 -1
1899 +11 +3 --6 -3 +1
1900 -I +4 -4 +5 +1
Here again quite a few irregularities are still evident. Years with too
much precipitation (+) and with too little precipitation (-) alternate repeated-
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 275
ly. However, some general trends are noticeable. The period of 1856-74
includes a number of very dry years which is indicated by an accumulation
of minus symbols; all years from 1877 to 1880 are too wet, and plus symbols
continue to outnumber the minus symbols slightly beyond that date. Still,
even these figures leave the impression that a relatively drier period in the
middle of the 70s was followed by a somewhat wetter period. This trend
becomes even clearer if we try to adjust the irregularities mathematically.
This can be achieved in different ways. In 1890, I chose the method of com-
bining five years each as shown in the following table for the entire area of
the German Empire. Figures indicate in percentages of the multi-year
average by how much the precipitation was too high or too low in the res-
pective five-year period.
Table 9.3. Variations of Rainfall in the Gennan
1821-25 -I 1866-70 +3
1826-30 -6* 1871-75 -6
1831-35 -6 1876-80 +9
1836-40 +2 1881-85 +3
1841-45 +3 1886-90 +l
1846-50 -I 1891-95 -2
1851-55 +8 1896-00 +2
1856-60 -7 1901-05
The lustra show an obvious pattern, 1816-35 are years with low
precipitation, 1836-55 with a lot of rain, 1856-75 with little rain, and from
1876-90 again with a lot of rain. Lately, I have been using a different
formula that I consider to be much better. I determine the averages of ten-
year periods (compare Table 9.2), e.g., 1851-60, 1852-61, 1853-62 etc. The
resulting figures depict a full decade and show the development of the
precipitation from year to year in averages of decades. Table 9.4 shows these
10-year averages next to the year which marks the centre of each time span
for which the mean value was established.
Table 9.4. Variations of Precipitation in River Regions of the Gennan Empire (in Percent).
Smoothed by 10-year Mean Values
Centre Year of the Weichsel-
10-Year Average Oder Elbe Lower Rhein UEEerRhein Mean
1855/56 -3 -I +2 -I -I
1856157 -3 -2 +1 -3 -2
1857/58 -5 -2 -I -3 -3
1858/59 -7 -3 -I -3 -4
1859/60 -9 -5 -5 -7 -7
1860/61 -12* -7* -6* -9* -9*
1861/62 -12 -6 -5 -9 -8
nico.stehr@zu.de
276 EDUARD BRUCKNER
Centre Year of the Weichsel-
10-Year Averal:le Oder Elbe Lower Rhein UEEerRhein Mean
1862/63 -6 -I ±O -4 -3
1863/64 -5 +1 +2 -I -I
1864/65 -4 +1 +1 -1 -1
1865/66 -4 ±O -1 -3 -2
1866/67 -5 -1 -1 -4 -3
1867/68 -2 -2 ±o -4 -2
1868/69 -2 -3 -1 -4 -3
1869170 -3 -4 ±O -2 -2
1870171 -2 -2 +3 +2 ±O
1871172 -2 -2 ±O +1 -1
1872173 -4 -2 -1 ±O -1
1873174 -3 -2 +3 +3 ±O
1874175 -2 -1 +4 +4 +1
1875176 ±O +1 +6 +6 +3
1876177 ±O +2 +6 +6 +3
1877178 +1 +5 +8 +9 +6
1878179 +2 +5 +9 +9 +7
1879/80 +4 +8 +9 +9 +8
1880/81 +5 +8 +7 +9 +7
1881182 +4 +6 +7 +9 +7
1882/83 +3 +3 +4 +6 +4
1883/84 +4 +3 +3 +3 +4
1884/85 +4 -3 +2 +3 +3
1885/86 +4 ±O ±O +1 +1
1886/87 +6 +1 +1 +2 +2
1887/88 +3 -1 -4 -3 -2
1888/89 +3 -3 -5 -4 -2
1889/90 +3 -3 -4 -3 -1
1890/91 +3 -2 -4 -3 -2
1891192 +4 -2 -4 -3 -1
1892/93 +5 ±O -4 -1 ±o
1893/94 +4 ±O -6 -2 -1
1894/95 +4 ±O -6 -2 -1
1895/96 +3 ±O -6 -2 -1
It becomes quite apparent now that all river regions of the German
Empire experienced a period of low precipitation from 1855-69, those in the
East up to 1874. This was followed by a period of high precipitation up to
1886, when rainfall decreased again. The decrease was much more pro-
nounced in the Rhein region than in the Weichsel and Oder regions where
precipitation continued to be too high up to 1900. I have not yet examined in
detail the precipitation variations beyond 1900.
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 277
Table 9.5. Variations of Precipitation in Central Gennany, Asia, and North America
Central Northern
Gennany Asia North America
by P. Schulz
Year {% adjusted} b~ Bruckner {cm adjusted}
Central
Nikolajew Upper Ohio Mississippi
Summer Fall Year Nertschinsk {Amur} Valle~ Valle~
1831132 -4 -4 -2 ±O
1832/33 -2 -5 -1 -4
1833/34 -3 -10 -4 -9
1834/35 -2 -15* -2 -10
1835/36 -10 -12 -6 -11*
1836137 -7 -14 -5 -8
1837/38 -10 -10 -6 -8
1838/39 -10 -9 -6* -5
1839/40 -18* -2 -6 -4
1840/41 -9 -1 -1 -2 -1
1841142 -7 -5 ±o +4 +5
1842/43 -6 -6 -3 +8 +4
1843/44 -8 +6 -1 +11 +7
1844/45 -11 +8 -2 +16 +10
1845/46 -4 +5 ±O +12 +12
1846/47 -8 +12 +2 +8 +16
1847/48 -2 +10 +5 +4 +17
1848/49 -2 +12 +4 +1 +14
1849/50 +2 +10 +5 -2 +10
1850/51 +2 +8 +2 -4 +6
1851152 +4 +8 +2 -2 +2
1852/53 +2 +2 +2 -5 ±O +27
1853/54 +7 -8 ±o -1 -2 +15
1854/55 +8 -9 ±o -2 -2 +7
1855/56 +10 -13 ±o -3 +1 +5
1856/57 +12 -18 -2 -6 +3 +6
1857/58 +17 -22 -1 -6 +4 +7
1858/59 +14 -21 -2 -10 +5 +8
1859/60 +11 -23 -5 -12* +4 +7
1860/61 +7 -24 -8 -10 +1 +4
1861/62 +2 -25* -9* -10 -12* -2 +2
1862/63 +1 -20 -5 -9 -12 ±o +1
1863/64 -4 -16 -3 -7 -12 +2 ±O
1864/65 -4 -10 -3 -5 -8 +3 -2
1865/66 -4 -8 -4 -4 -7 +5 -2
1866/67 -4 -13 -5 -4 -10 +4 ±O
1867/68 -10 -6 -6 +1 -11 ±O -1
1868/69 -9 -6 -7 -1 -12 -5 -2
1869/70 -13 -8 -7 +3 -10 -9 -3
1870171 -12 +1 -4 +1 -11 * -5
1871172 -15* +5 -4 +1 }-11 -11 -7*
1872173 -12 +5 -5 ±o -8 -5
nico.stehr@zu.de
278 EDUARD BRUCKNER
Central Northern
Gennany Asia North America
byP. Schulz
Year {% adjusted} b~ Bruckner {cm adjusted}
Central
Nikolajew Upper Ohio Mississippi
Summer Fall Year Nertschinsk {Amur} Valle~ Valle~
1873174 -8 +2 -5 +2 +5 -5 -I
1874175 -4 ±O -3 -3 +5 -3 +3
1875176 -5 +4 -2 -2 +10 -3 +6
1876177 -6 +8 -I +2 +9 -3 +8
1877178 +'4 +10 +3 +3 +13 -2 +9
1878179 +6 +14 +4 +6 +4 +2 +10
1879/80 +9 +19 +8 +12 +4 +7 +10
1880/81 +9 +16 +6 +14 +3 +12 +11
1881/82 +13 +12 +7 +12 +1 +13 +II
1882/83 +10 +11 +6 +6 +1 +10 +9
1883/84 +10 +16 +6 +1 +I +5 +5
1884/85 +8 +17 +6 -3 +3 +3 +1
1885/86 +9 +14 +5 -6 +3 -3 -2
1886/87 +9 +12 +6 -6 +5 -2 -6
1887/88 -1 +2 ±O -2 +5 +2 -7
1888/89 -4 +4 ±O -4 +6 +4 -7
1889/90 to +5 ±O -4 +2 +4 -6
1890/91 -1 +4 +2 -1 ±O +I -5
1891/92 +I +6 +2 -2 -1 -4 -7
1892/93 +2 +7 +2 ±O -2 -10 -8
1893/94 +1 +2 +2 +3 -1 -13* -10
1894/95 to ,to +2 +2 -2 -11 -11
1895/96 -1 -1 +2 +1 -3 -8 -11*
1896/97 -5 +5 +2 -2 -8 -3 -10
Upon my recommendation, Dr. Paul Schulz in his thesis 448 determined the
average precipitation of each meteorological season exactly as I had done for
the annual precipitation. His figures for central Northern Germany are
included in Table 9.5 according to year, summer, and fall; they have been
adjusted by forming ten-year averages and are expressed in percentages of
the multi-year average.
It became evident that the figures for summer and fall as well as the
annual averages indicate an arid period in the 1830s and up to the beginning
of the 40s, then a wet period in the 1840s and the first half of the 1850s
which, however, has a late start in the summer season, followed by a dry
period in the 1860s up into the first half of the 1870s, and then again a wet
448 Klimaschwankungen im mittleren Norddeutschland und ihr EinflufJ auf die Ernteertriige
[Variability of the Climate in central Northern Germany and its Influence on the Harvest].
Inaugural-Dissertation, Halle a. p., 1907.
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 279
period which according to Schulz' figures does not fade out until towards the
end of the century. In contrast, winter and spring (omitted here) show
irregularities.
In order to demonstrate the simultaneous occurrence of these variations
in most parts of the global land masses, I have added to Table 9.5 the results
for the upper Ohio- and central Mississippi-valley in North America, as well
as for Nertschinsk (smelter plant) and Nikolajewsk on the Amur in Asia. 449 It
is obvious how completely coincidental the variations are, as shown in the
following table for each continent (annual averages, percentages).
Table 9.6. Variabili~ ofPreciEitation for Each Continent
Five-year North Central and Globally
SEan EuroEe Asia America South America Australia {Land Masses2
1806/10 +3 +3
1811/15 iO -7 -3
1816/20 -I +22 -4 +6
1821/25 -2 iO +4 +1
1826/30 -I +14 -4 +3
1831/35 -10* -7 -12* -6 -8*
1836/40 -1 -14* -7 -19* -5
1841/45 +4 +9 -3 -11 -11 +1
1846/50 +1 +13 +16 +8 +17 +3
1851/55 +4 +2 +3 +16 +16 +1
1856/60 -4 -9 -8 +5 +2 -4
1861/65 -10* -13* -10* -11 -6* -5*
1866170 iO -9 +7 -12* +10 -1
1871175 iO +8 -\ -9 +14 +2
1876/80 +10 +20 +4 +7 -1 +7
1881/85 +6 +23 +13 +10 -19 +6
It is evident that the dry period of the 30s and of the 60s occurred in all
the continents. So does the wet period of 1850, as well as the one around
1880, which peaks in Australia five years earlier than in the other continents.
Based on the averages for the global land masses, I was able to determine
more precisely that the time from 1826 to 1840 was dry, 1840 to 1854 wet,
1855 to 1871 dry, 1872 to 1887 wet, and since 1888, fairly dry again. In this
tabulation the general pattern emerged (compare Table 9.1 and 9.6) that the
cool and the wet periods coincided, and so did the warm and the dry periods.
The variations of precipitation are quite pronounced, especially so in the
largely continental regions. In East England they are 16%, in Northern
Germany 20%, in Southern Russia 24%, in Southeast Russia 35%, in West
449 These numbers from my essay of 1902 are not expressed in percentage of the long-term
average but state rainfall height in cm. The adjustment of a ten-year mean has not been
made either but through a five year mean. The second figure for the year of each row show
a five-year average in relation to precipitation figures.
nico.stehr@zu.de
280 EDUARD BRUCKNER
Siberia 86%, in East Siberia 47% of the multi-year average. They are also
significant in the interior of the United States of America (36%).
It is noteworthy that the islands in the middle of the ocean, such as
Iceland and Ireland, but also parts of the oceanic coast lines show the oppo-
site pattern, i.e., in those periods when the average precipitation is low for
the global land masses, it is high for the ocean regions, and vice versa. This
is easily explained if air pressure is taken into account. High temperature~
accompany the dry periods of the continental areas, which means that in
middle and higher latitudes the air pressure over oceanic areas is slightly
lower than during the wet and cool periods of the continental areas. If,
however, the air pressure over the ocean is relatively lower than on land
during warm periods, which has been established in particular for the region
of the North Atlantic Ocean on the one hand and Central Europe on the
other, the transfer of wet ocean air to the continent is more difficult, whereas
it is easier when the air pressure over the ocean is not quite as low, as is the
case during cool periods.
Meteorological observations have been around for slightly more than a
hundred years. Climate variations, however, can be traced back much further
by other methods. It has been shown, for instance, that the precipitation fluc-
tuations of the 19th century are quite poignantly reflected by the water level
fluctuations of rivers. Moreover, the temperature variations accompanying
the changes in precipitation affect the timing of the grape harvest in the wine
regions of Europe as well as the freezing and melting of the rivers in Russia
and Siberia. Regularly kept records about the harvesting of grapes in France
go back to the year 1391 and those about the melting and freezing of
Russia's rivers to the year 1556. Based on these data, I was able to trace the
climate variations back over five centuries. In addition it was possible to
establish the average time period of a climate variation; it is about 35 years,
i.e., the length of time between the peaks of one wet and cool period and the
next is about 35 years on the average. Occasionally, however, this time
period varies in length so that periods may last 40 or even 45, and others 30
or only 25 years. This irregularity is not surprising: meteorological cycles
never occur with mathematical precision. For example, the length of time
which lies between the ten hottest days of one year and those of the next is
not always the same depending on whether these hottest days fall into the
month of June, July or August.
My research results on climate variability have been confirmed by a
number of scientists. J. Hann was the first meteorologist among my
contemporaries to do so; based on extremely thorough research methods
involving a number of stations in the region of the Alps, he was able to
provide strong evidence of climate variations for the alpine countries. Other
scientists, like B. Kremser, are more reluctant. I would like to stress that in
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 281
long-tenn mean values, as demonstrated above using ten-year averages, the
climate variations are clearly evident for almost all continents. However, if
individual years alone are considered, the random irregularities which the
precipitation is subject to from year to year, partially obscure the large
variations.
Climate variations affect human life in many ways. I have commented on
this in detail in previous essays. Using long-tenn averages I have demonstra-
ted in 1895 that agricultural crops in particular are influenced by climate
variations, which Dr. Paul Schulz confinned in his dissertation in 1907. This
is shown in the following table. The figures up to 1890 quoted from my
article in the Geographical Journal, Volume I are average values for the eight
old Prussian provinces; those from 1891 onward (in brackets) are taken from
Dr. Schulz' dissertation and refer to the provinces Brandenburg and Saxony.
Yields are expressed in percentages of an average crop and their deviations
from 100. For example, the figure -19 indicates that on the average the
wheat crop of 1881-85 was 19% below average. The three bottom rows
combine the years according to wet and dry periods.
Table 9.7. V ariabili~ of CroE Yields in Prussia
Five-~ear Period PreciEitation Wheat R~e Barle~ Oats
1846/50 -I +7 +10 +4 -2
1851155 +8 -2 -4 +3 +2
1856160 -7 +2 +7 -10 -9
1861/65 -9 +2 +3 +7 +8
1866/70 +3 -2 -2 -3 -3
1871/75 -6 +2 -I -1 -2
1876/80 +9 -11 -14 -9 -8
1881/85 +3 -19 -23 -19 -22
1886190 +1 -13 -22 -18 -12
1891195 -2 (+6) (-15) (-3) (-2)
1896/1900 +2 (+28) (-12) (+16) (+27)
1856175 -5 +1 +2 -2 -2
1876/90 +4 -14 -20 -15 -14
189111900 ±O (+17) (-16) (+6) (+13)
During the two fairly arid decades 1856 to 1875 the wheat, rye, barley,
and oats yields in Prussia were relatively good; in contrast the wet 15 years
1876 to 1890 had poor yields-according to assessed average crop.
Particularly the wheat, barley and oats yields increased under dryer weather
conditions.
Also grain prices reflect the impact of climate variability. These prices
can be traced back further than data about the outcome of crops. For
Gennany, e.g., according to records of wheat prices for Miinchen, it clearly
indicates that the wet periods of the 18th Century and in the first half of the
19th Century show high prices and the dry periods low prices. The same
nico.stehr@zu.de
282 EDUARD BRUCKNER
goes for England and France. Crops in these regions suffer more likely
because of too much rain than of too much dryness. The reverse is the case
in Russia. Here arid years are characterised by high prices, wet periods by
low prices. However, the impact of climatic variability on wheat prices in
European regions near the ocean is less noticeable since the middle of the
past century. The outcome of the home crop is no longer the main guideline
for pricing but rather the one of the world's major grain producers: Russia
and the United States.
In Russia and the United States good harvests occur during wet periods;
on the other hand, poor harvests occur in the oceanic regions like England,
especially in Western Europe, and in most of Germany. Consequently,
during wet periods in oceanic regions an increased demand for imports is
created, whereas increased export opportunity of continental [inner Euro-
pean] grain suppliers occurs as well. Wet periods stimulate an international
wheat trade. As a result, this triggers price reductions in regions with already
bad harvests, unless prevented by protective duties. The result is a severe
distress in agriculture because in addition to poor harvests, prices are low.
This situation becomes different during arid periods. In oceanic regions,
harvests are better than before, and continental grain suppliers suffer losses
because of dryness. Therefore, demand of the first is lower and export
capability of the latter has also diminished. An arid period will therefore
coincide with a period of diminished international grain trade, and pricing in
oceanic regions will respond more positively to their own harvest conditions.
Such an impact of climate variability may not fully affect market
conditions. Political and economic events naturally also playa major role.
Nonetheless, our reports show that the influence of climate variability on
grain prices cannot be cancelled out completely by human interference.
Finally, I would like to raise the question if weather predictions for
individual future years are possibly based on climate variability. I would like
to point out, and as undoubtedly seen in the above, this is not feasible. From
the past climate variability curve range we can only conclude that a certain
future long-term period (10 to 15 years) could be wetter or dryer on the
average. Within a wet period more wet years than dry years will occur, but
dry years will not be lacking completely as shown in Table 9.2. On the other
hand, wet years will also occasionally occur during arid periods. Based on
climate variability for single years, it is not possible to predict weather
conditions for agriculture under such circumstances. If we look at the
climate variability pattern of the past 200 years, the conclusion that this will
also occur accordingly in future cannot be fully ignored. Around 1880, land
masses on earth experienced a wet period. Precipitation has more or less
decreased ever since. Based on the average length of climate variability, the
centred maximum of the next arid period should be around 1900. However,
nico.stehr@zu.de
ABOUT CLIMATE VARIABILITY 283
an exact prediction of the situation will only be possible when precipitation
observations of additional ten years are available. Still, I believe to be able to
predict that we are approaching a wet period again.
nico.stehr@zu.de
Chapter 10
Climate Variability and Mass Migration *
We may look upon today's economy as a huge mechanism. Multiple factors
interlink like wheels of a clockwork. They all have an impact on the end
result and through probability studies alone can the significance of one or the
other factor be isolated. It is obvious and quite evident that the natural envi-
ronment on earth affects the development of human life and culture. Yet one
generally tends to attribute changes in the economic conditions of large or
small regions to man-made causes. Nature's factors are assumed to be con-
stant, and only because man's acquired knowledge changes over time is their
effect different at different times. Overall this is quite correct-but not
always. In some instances natural phenomena can be contributing factors of
considerable significance in economic processes. Their significance is
masked by the fact that the mechanism of economic life is so very complica-
ted. The effect of these natural causes is enhanced by political and social
factors. Permit me to concentrate on such combined influences at this point.
One of the most striking aspects in the history of the XIX Century is the
mass immigration from Europe to the New World. From 1805 to 1911, no
less than 28 112 million people left old Europe and emigrated to the United
States of America. The rapid settlement in vast regions of the United States
was the immediate consequence of this emigration from Europe. The boun-
dary of the "oecumene", as F. Ratzel called the populated land, advanced
further west each year. The massive number of people mobilised during this
time is certainly not smaller but considerably larger than the number of
people who mass migrated in the early Middle Ages. Emigration to America
is the largest mass migration of all times. Economic conditions in some parts
of Europe had suffered because of overpopulation and were the reason for
emigration, while especially the sparsely populated western part of the
United States, where as a result wages were high, was a strong incentive for
immigration. Though these undoubtedly are the key elements setting this
• Klimaschwankungen und Volkerwanderungen, Vortrag Kaiserliche Akademie der
Wissenschaften, Wien 1912.
285
nico.stehr@zu.de
286 EDUARD BRUCKNER
modem mass migration in motion, there are a number of aspects indicating
that other than human factors were involved, sometimes precipitating migra-
tion, sometimes slowing it down.
Economic conditions in the United States as in major parts of Europe are
predominantly dependent on harvest conditions. Years of poor harvests, par-
ticularly when suffered in succession, stimulate emigration from the affected
European regions. If at the same time in another area harvests are good and
the economy is prospering, these factors are an additional incentive to leave
the country. Regarding the turnout of crops, there is in fact this polarity
between Western and Central Europe on the one hand and the United States
on the other.
There would be no agriculture without water, and then again no agricul-
ture with too much water! This dependency is obvious when one looks at the
extent of agricultural land around the globe and examines the causes of poor
harvests. In the regions with little rainfall, droughts and poor harvests go
hand in hand; in regions with excessive rainfall, poor harvests occur mainly
during wet years. It is of course not the absolute amount of rain that makes
the difference, but the ratio between rain and evaporation. The same amount
of precipitation that drowns the crop in a cold climate may barely be
sufficient for growing grain in a hot climate. In contrast, low rainfall that in a
hot country may cause drought due to high evaporation is frequently entirely
sufficient in a cool climate. In Europe, conditions in the wet countries near
the North Atlantic Ocean, as for instance Norway, Denmark, Ireland, and
Great Britain in particular, but also Sweden and Central Europe are the exact
opposite to those in the dry interior of the continent. Southern Russia on the
one hand and Great Britain on the other are the extremes in this regard.
Drought is almost always the cause of crop failure in Southern Russia, while
in England the numerous crop failures at the end of the thirties, during the
forties, at the beginning of the fifties, and again in the seventies and eighties
occur in excessively wet years. Southern Europe and most of the tropics, at
least as far as they are grain-growing countries, have similar conditions as
those prevailing in Southern Russia. Years of famine in India coincide with
dry years. The same pattern is clearly evident in the United States of North
America where crops increase and decrease with the amount of precipitation.
As a result we may conclude that the simultaneous occurrence of a number
of wet years in Western and Central Europe and in the United States of
North America must affect the crops in these areas in a completely different
way. Western and Central Europe will suffer from poor harvests, regions of
the United States will have good ones. These are the conditions which
encourage emigration from central and western Europe to the United States.
The situation is different, when the two regions simultaneously go through a
number of dry years. In that case, poor harvests make it less attractive to
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRA nON 287
immigrate to the United States while good harvests at home persuade people
to stay.
If wet and dry years would change purely by chance in place and time,
poor or good harvests could not succeed one another over a number of years;
they would show the same random pattern. That is in fact not the case. My
own research all around the globe has shown that wet and dry years often
occur in groups, so that climatic conditions are not completely constant. The
climate varies instead, oscillating around a middle range. My results have
been confirmed by various parties, above all by Julius Hann.
Climate variability consists of multi-year fluctuations of temperature, air
pressure and rainfall, all of which occur simultaneously all around the globe.
Temperature is the one element on which all the others depend. Temperature
variations are common to nearly all of the countries on earth. All of them go
through cold periods and warm periods at the same time. For example, the
five-year periods of 1806-20 have on average been too cold all over the
globe, those of 1821-35 too warm, 1836-50 again too cold, 1851-75 too
warm, 1876-90 again slightly too cold.
Temperature variations have an impact on the distribution of atmospheric
air pressure. In Europe as well as in North America the transfer of moist
oceanic air from ocean to land seems more difficult during warm periods,
and easier during cold periods. Consequently, on both continents warm
periods tend to be, at the same time, dry periods, whereas cold periods show
a higher amount of rainfall. The cold periods around 1815, 1850 and 1880
for instance were wet periods in both Europe and North America, while the
warm periods around 1830 and 1860 show very little precipitation. Rainfall
has slightly decreased since 1890; this is clearly evident in the United States
of America but also in Central and Western Europe.
Over the past two centuries, the centres of cold and, for the land masses,
wet periods fall into the years 1705, 1740, 1775, 1815, 1850, and 1880;
centres of warm and, for the land masses, dry periods into the years 1720,
1760, 1790, 1830, and 1860. Hence, one climate variation takes an average
of approximately 35 years from one maximum to the other. The centre of the
latest warm and dry period should occur around 1900. This however is a
preliminary estimate only because the climatological data from all over the
world have not yet been interpreted in this regard.
This climate variability clearly affects crops even if in a different sense
for different areas. Some examples may demonstrate this more clearly.
The variability of crop yields is very pronounced in Prussia. The top
curve in the following graphic (Figure 10.1) shows the rainfall fluctuations
in periods of 5 years each.450
450 Smoothed according to the formula (a +2b+c):4 =smoothed value of b.
nico.stehr@zu.de
288 EDUARD BRUCKNER
1848 53 58 63 68 73 78 83 88 91
.,~
"'
4%
R /
'" 1/ .............
0%
~ /'
"- . /
--
-4%
wz I"""-
......... o
""
RO
i"o.. ./
.......... ~ ./ o
"" -- ~
1848 53 58 63 68 73 78 83 88 91
" ~
Figure 10.1. Variability of rainfall in relation to the grain crop in Prussia. The grain crop
(WZ =wheat crop, RO =rye crop) is in percentages of an average crop, i.e., in deviation from
a multi-year mean (1 increment = 5% deviation); rainfall (R) is also in deviations (%) from
the multi-year mean (I increment = 4%).
1853 58 63 68 73 78 83 88
RO
~
/
f """"
~
'\. , ..........
ObI
'"
~
'I
"'"
.til
/ "-
VJ P"""
~ 1%
'\.
",
~
R
/ .,
["'0.... ~
/ , /
J
16 Bush
wz J
/
~
o
""
R
~ ./
/
/
~
~ 1/ 0%
" ~
1853 58 63 68 73 78 83 88
Figure 10.2. Variability of rainfall and of the increase in rye export from Russia and of
wheat crop in Ohio. One scale mark equals 2% for the rainfall (R), 1.5 million hectolitres for
the increase in the export of rye from Russia (RO) and 0.8 bushel per acre for the wheat crop
in Ohio (WZ). The variation in the increase of the rye export from Russia was determined as
follows: A straight line was drawn, according to the method of least squares, through the
recorded quantities exported from 1851 to 1890, and for each five-year period, the differences
between this figure and those actually observed were then depicted.
Around 1860 rainfall was low, around 1880 it reached a maximum.
Because in Prussia harvests suffer more frequently from too much rain than
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRATION 289
from too little, yields were high during the dry period around 1860, and
much lower during the wet period in the early eighties. The curves of the
crop yields 451 are the exact mirror image of the rainfall curve. The same
pattern applies to the crop situation in the entire area of Western Europe and
also to Great Britain in particular. The pattern is reversed, however, in the
United States of North America and in Russia, as demonstrated in Figure
10.2. Again, in the bottom curve for Ohio we recognize a minimum of rain-
175863 68 73 78 83 88 83 981803 08 13 18 23 28 33 38 43 48
34M
30M 17Fr
26M 15Fr
+41 13Fr
+21 0%
Ot
-21 10%
175863 68 73 78 83 88 83 98180308 13 18 23 28 33 38 43 48
Figure 10.5. Variability of rainfall and wheat prices in Southern Germany and
Switzerland. WM = wheat price, in Marks, of a Bavarian bushel of wheat in Munchen (1
scale mark = 2 Marks): WZ = price, in Swiss Francs, of 50 kilograms of wheat in Ziirich (1 =
1 SF); R = rainfall in southern Germany (l = 3% deviation from the multi-year mean); E =
date of the vine harvest in southern Germany and Switzerland which varies proportionately
with the rainfall, being late in moist and early in dry years (1 = 1 day; the multi-year mean is
marked Ot).
fall around the year 1860 and a maximum at the beginning of the eighties, as
was the case in Prussia. But since the U.S. crops are mainly affected by
droughts, the dry period around 1860 results in poor harvests, the wet period
at the beginning of the eighties in good ones. For Russia harvest statistics
dating further back in time are missing. I have used the increase in rye export
from Russia as an indicator of harvest conditions. The bottom curve again
reflects variability of precipitation with a minimum around 1860 and a
maximum around 1880. The obvious parallels between the two curves is so
effective that the crop yields fall and rise with the rain.
Harvest statistics do not go back very far. But at a time when there was
no world trade in grain, the outcome of a country's harvest also determined
grain prices at home. To a certain extent the grain price is, therefore, an
indicator for the quality of the crop yield in a region. In times of poor
harvests grain is expensive, in times of good harvest it is cheap. Figure 10.3
shows the close correlation between wheat prices and harvest conditions and
also the rainfall in southern Germany and Switzerland. Again slightly
451 Smoothed as before.
nico.stehr@zu.de
290 EDUARD BRUCKNER
smoothed five-year averages are used as a basis. Rainfall shows pronounced
maxima around 1770, 1810, and 1850, minima around 1785, 1825, and
1860. Wheat prices in Miinchen and Zurich fluctuate accordingly, as can be
read from the curves. This obvious correlation is missing from the middle of
the last century on because Switzerland started to import grain from Russia
and the Balkan states. Grain prices in Zurich were no longer determined by
harvest conditions at home, but by the conditions of the harvest in those
continental countries.
England has data about rainfall and wheat prices dating back to the
beginning of the 18th Century. Figure 11.4 illustrates this.452 During the wet
periods around 1713 and 1768 up to 1773 grain prices were high and
harvests poor; in the dry period in between, however, the price was low. The
similarities of the curves comes as a surprise, yet it disappears completely
towards the end of the 18th Century. At this time political events, such as the
implementation of the continental blockade are responsible. Later the world
price for grain evolves, a criterion even for England's grain prices, which is
gradually abolishing grain duties.
1703 08 13 18 23 28 33 38 43 48 53 58 63 68 73 78 83
I I I i i i i I
+10%
,,
"-
,,
...... -~ +5%
.....
R i-"
" - 0%
-
.....
'"""~ ..... " ....... '-
1'0..
I " 40sh
,. '" ""' - '"
W
~ L
.....
"
170308 13 18 23 28 33 38 43 48 53 58 63 68 73 78 83
Figure 10.6. Variability of rainfall and wheat prices in England. Rainfall (R) is indicated
in deviations from the multi-year mean (percentages) (I increment 2.5%), the annual =
average wheat price (W) in Shillings per Imperial Quarter (I increment = 2 sh.).
In summary, we can state that as a result of the wet years around 1815,
1850 and 1880, harvests in the oceanic regions of Europe, including central
Europe, were poor, whereas in the United States and Russia they were good.
The situation was reversed during the dry periods around 1830, 1860, and at
the end of the century. It is important to note that we are not dealing with
single occurrences of good or poor harvests but rather with a number of
harvests of the same quality grouping around those years.
452 Again smoothed as before.
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRA nON 291
That it is mainly the fanning population which suffers from poor harvests
is obvious. On the other hand, it is a known fact that the large number of
emigrants to the United States came from the fanning population of Central
and Western Europe. Therefore, the question arises whether perhaps climatic
variability by way of varying crop yields is also reflected in the emigration
figures to the United States.
In fact U.S. immigration statistics are very detailed and date back a long
time. The following graph (Figure 10.5) gives an overview. The top curve
shows the rainfall rate in the United States, the second one the rainfall in
Western Europe. Rainfall maxima occur around 1850 and 1880, the minima
around 1860. The next curve shows the U.S. immigration figures.
The first thing that catches the eye is the enonnous increase in the num-
ber of immigrants. Yet it is not a steady increase. Periods of rapid increase
alternate with times when immigration figures remain the same or even
decline. In the years 1821-35 immigration was relatively low, after which it
began to increase rapidly. Numbers continued to grow until the middle of the
fifties. It coincides with the weather period which brought a lot of precipita-
tion to both Europe and in the United States, resulting in poor harvests in
Western Europe and in good ones in America. From approximately 1855
until the middle of the sixties immigration figures fall by more than half.
This coincides with the arid period leading to good harvests in oceanic
Europe and bad harvests in the United States. The American Civil War was
certainly a contributing factor at the beginning of the sixties, but the flow of
immigrants had already started to ebb 6 years earlier. As a consequence of
the wet period peaking at the beginning of the 80s, immigration rapidly
increases peaking at more than 400,000 people per year between 1880 and
1893 and reaching a maximum of 780,000 people in 1882. This is followed
by a period of low immigration from 1894-1900, which then however is
followed again by a period of rapid increase, leading to an unprecedented
high of 1,285,000 immigrants in 1907. This last rapid increase coincides
with an arid weather period and appears to contradict our claims. This,
however, is not at all the case as we shall see when we take a closer look at
the demographic factor of the immigration movement.
nico.stehr@zu.de
292 EDUARD BRUCKNER
1833 43 53 63 73 83 93 1903
0/0
+10
Rain j \ United States /' ~
\,
+5
o
I \ V
/
, V
I\,
"
I ... "- /
-5
~
)
,
+5
o
I ~
"." 1'00.. I ~ J 350
-5
~ ~W-Europe
Rain ~
r--... ,/ """
J\
~ I 300
I 250
~
"- I
I "r--... II
,f 200
150
j ~ / \~
-V
....... ........ 1( 100
50
i--"" Total Immigration into the United States
80 /~
60 J \ ,..-.. J
~
\
40 / \ V 1\ / \
"'" I"V
20
".
~
Bri~ish ~~igr~tion, int~ "
the, Uni,ted ~tat~s
1833 43 53 63 73 83 93 1903
Figure 10.7. Variability of rainfall in the United States and Western Europe and the total
number of immigrants to the U.S. and from Britain. The curves are based on five-year
totals (non-smoothed). The number of immigrants is given in 10,000s, rainfall in deviations
(percentages) from the multi-year mean.
Until the end of the past century the majority of emigrants came from
Great Britain and Ireland as well as the German Empire, that is from the
oceanic regions of Europe. It is a striking fact that emigration from these
oceanic regions increases during wet periods and decreases considerably
during dry periods. It is high in the wet period around 1850 and 1880, and
considerably lower in the dry period around 1860. If, for example, the data
about emigrants from the German Empire to America and about the
precipitation in the German Empire are presented in a graph the resulting
curves show a striking similarity (compare Figure 10.6). Even during the
secondary maximum of precipitation in the second half of the sixties
emigration does increase rapidly. After a slight delay the secondary
minimum of rainfall during the first half of the seventies is followed by a
minimum number of emigrants. The number of extremely wet years in the
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRATION 293
late seventies and early eighties which are particularly pronounced in the
German Empire, resulted in the highest number of emigrants to America
which this country had ever reached. Several times the emigration curve is
lagging behind the rainfall curve by five years. This is not surprising,
because of the logical delay between cause and effect.
For Great Britain the situation is similar (Figure 10.5, bottom curve).
Again emigration to the United States increases and decreases with the
rainfall. And again, an occasional delay of a few years can be observed.
A closer look at each year reveals that emigration to America from Great
Britain reached a maximum in 1851, from Germany in 1854, and a second
time in 1888 and 1881 respectively. During the past century the largest
number of immigrants to the United States was reached in the years 1854
and 1881. All these years fall into wet periods.
1833 38 43 48 53 58 63 68 73 78 83 88 93 98 1903
- , .....
I i I I i i
+10% r-Rainfall
+5%
0% """ '/
""- , ~
.....
---
-5% -' '- I
100
- -
80
60
40
20 ~
. / I" ./ .......
1.1'" ........
f
I
" ~
r"..
o
-Elmigfati~n
1833 38 43 48 53 58 63 68 73 78 83 88 93 98 1903
Figure 10.B. Variations of rainfall and emigration from the German Empire to the
United States. The curves are based on five-year totals (non-smoothed). Rainfall is the
deviation (percentages) from the multi-year mean. Emigration to the United States is in
10,000s.
As we have seen, starting in 1900 the number of immigrants to the
United States rises sharply. Emigration, however, from oceanic Europe does
not reflect this development. The rapid increase is instead caused by a factor
that heretofore had played only a minor role: massive emigration from
Russia, Galicia, Hungary, and Italy. This conforms again with the climate
variations. The same dry period that produced good harvests in the oceanic
regions of Europe caused poor harvests in Russia. In addition, the sad state
of Russian politics, the persecution of Jews, the Japanese war, and the
revolution encouraged emigration. In contrast, emigration from Germany to
the United States remained very low during that entire dry period beginning
in 1893 and barely reached half of the numbers that had emigrated in the
eighties and at the beginning of the nineties. No doubt, the enormous
industrial boom had something to do with it, keeping large numbers of the
nico.stehr@zu.de
294 EDUARD BRUCKNER
German labour force in the country which otherwise might have emigrated.
Nonetheless, the arid period and its good harvests should not be overlooked.
In the same way as the impact of climate variations manifests itself in the
massive emigration from Europe to the United States, it also does so in the
regional settlement of the United States.
If we compare the census data of several years regarding the distribution
of the population, we find a prevailing pattern. The majority of the
population is slowly moving westward. This trend led to the rapid settlement
of the far West of the United States, especially in the seventies and eighties
of the past centuries. Population in the states West of the Mississippi
increased by 79% from 1870 to 1880, then from 1880 to 1890 by 71 %.
These are average values; in some of the states this increase was many times
higher. In contrast, from 1890 to 1900 the population increased by not more
than 32%. These western states are obviously going through a decline at that
time which the Atlantic and Central States are not experiencing.
The reasons for this decline lies, no doubt, in the fact that the Western
States in the short period from 1870 to 1890 had increased their population
to capacity through immigration. What has remained unnoticed and, in my
opinion, is certainly not a coincidence, is the fact that this settlement rush in
the far Western United States occurred during a period of high precipitation.
Statistics about precipitation in the United States show a series of
predominantly wet years in the interior of the United States at the beginning
of the seventies until the mid eighties. This was followed by a very severe
dry period in the nineties. I am inclined to attribute the slowdown of the
population growth in the Western States from 1890 to 1900 at least in part to
that arid period.
Moreover, with the beginning of the series of dry years, people in the
Western States were forced to abandon those regions that had been cultiva-
ted during the previous wet period. A sufficient water supply was available
for agricultural use at that time, which now during the dry period of the
nineties of the 19th Century and at the beginning of the 20th Century was no
longer available. Density of population charts as part of the census data
shows that from 1890-1900 the settled area of the western states shrank
considerably. On the basis that areas with at least two people per Imperial
square mile are defined as settlements, New Mexico has a decrease of
around 61 % from 1890 to 1900, Nevada of 86%, Idaho of 58%, Kansas of
12%, etc. North Dakota is an exception. The total populated area of all 13
western states combined went down by 15%, or about 25,000 sq. km from
1890 to 1900. The area that people have abandoned is about four-fifths of
the size of Austria.
What causes this population backflow from those affected Western
regions? Again, we cannot blame a single event but rather a complex system
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRATION 295
of several events. Since the settlement declined mainly in already sparsely
populated areas, a reduction in the agriculturally used land had to follow.
What causes such a depletion? Reasons vary from state to state. In states
where farming is closely connected to the mining industry, serving the
demands of the local mining communities, a reduction in mining naturally
results in a reduction in farming. That is indeed the case in Nevada and
perhaps in some parts of Colorado. The situation is different in states where
farming plays a major role, such as Kansas, Wyoming and Nebraska. Here
the decline in farming can undoubtedly be attributed to the dry period that is
evident since the second half of the eighties and particularly since the
beginning of the nineties. Abandoned farms are proof of this decline in these
states. Yet the overall population has not decreased. People moved from the
affected farmland to more fertile areas or into the cities where industry and
trade were booming. The number of small country towns, which had mainly
developed as centres for these farming communities, has now gone down
considerably. If the foremost agricultural state of North Dakota continues to
show strong population increases from 1890 to 1900 and the size of South
Dakota's population does at least stay the same, it can be attributed to the
fact that these states were settled relatively late and were still able to accept
additional settlers at the onset of the dry period. In Nebraska and Kansas as
well as in all the other states with similar conditions vast areas had been
converted into farmland during the wet period, which in the following arid
period did not produce. These are examples of the correlation between
climate variations and the varying sizes of agriculturally used land. Utah is
another example. This state experienced a rapid increase in population
during the seventies. The size of the farmland grew enormously while, at the
same time, the water level of the Great Salt Lake rose. The Mormons
attributed the rising lake level to the increase of agricultural land in the area
and assumed this was the cause of higher precipitation. But the arid period
starting around the middle of the eighties caused the lake's water level to fall
again and proved that the fluctuations of rainfall were unrelated to man's
cultivation efforts. Instead they occur in connection with the global
phenomenon of climate variability.
Let us repeat! In the oceanic regions of Europe, poor harvests as a
consequence of wet periods led to a large wave of emigration. At the same
time, these same wet periods enable the cultivation of vast areas for
agriculture in the continental regions of the United States. At this point the
dry period sets in. Emigration becomes less attractive in Europe's oceanic
regions and in the continental states of America people are driven back from
large areas of previously cultivable land. Mass migration, which we see
happening as a result of climate variations, is indeed an enormous phenome-
non. Though climate change is not the only cause for mobilising this stream
nico.stehr@zu.de
296 EDUARD BRUCKNER
of people, its strong influence is apparent in the ebb and flow of the migra-
tion movement. Political and economic crises affecting industry, but which
also to some degree depend on harvest conditions, to some degree cannot
disguise this influence. This is always clearly manifested in the long-term
averages.
The enormous phenomenon of mass emigration from the Old World to
the New World is generally not called a mass migration, although the
number of people mobilised in the course of this event was much larger than
those in history's many mass migrations from Central Asia to Europe or
China. No doubt, the process of this modern mass migration is more
complex and its causes are more difficult to isolate. But we have the great
advantage that we can trace it statistically and that in our inquiry into the
influence of climate variations we can rely on statistics about the amount of
precipitation as well as about harvest yields and grain prices. Such data is
lacking for an inquiry into the causes of historical mass migrations. For
instance, we know nothing about the conditions in the home countries of
those hordes of riders from inner Asia, and are unable to identify the cause
for their migration. We can only assume that again climatic variations played
a role. Aurel Stein has introduced us to the remnants of buildings in the
middle of the sands of the Tarim basin where nowadays the desert makes
human settlements impossible. Only a small part of his extensive research
material found in these desert cities has so far been published, but already
the outline of those regions' history of settlement is coming to light. It has
been found that a number of settlements from the beginning of our time have
disappeared by the end of the 3rd Century. It cannot be coincidental that
only a few decades later a massive number of migrating people reached
Central and Western Europe. This does not mean that those who reached
Europe were the occupants of the settlements now in ruins. Rather the
demise of these settlements seems to be indicative of deteriorating climatic
conditions, which made the steppe of central Asia uninhabitable.
Climate deterioration in Asia also preceded the invasion of the Mongols.
We have evidence of this climate change: According to buildings along its
shoreline whose age could be determined, we know that the Caspian Sea had
reached its lowest water level in the 12th Century, which was unprecedented
and was never again reached thereafter. A low water level of this magnitude
could only have been triggered by a long arid period. This aridity cannot
have been a local occurrence but must have affected the entire Volga region
as well as large regions of central Asia. I am inclined to see this drought as
the incentive for the invasion of Europe, India and China by the Mongolian
hordes. As a consequence, many aspects support the theory that even the
mass migrations of the past were caused by strong climate variations.
Huntington even speaks of the pulse of Asia. In the rhythmic intervals of a
nico.stehr@zu.de
CLIMATE VARIABILITY AND MASS MIGRATION 297
pulse beat, Asia, from time to time and in accordance with the fluctuations of
the climate, releases waves of people into the peripheral regions of the Old
World. They surge threateningly against the borders of Europe's old
cultures; some disappear in these cultures, others are rejected. It should not
be overlooked that political conditions play a role here as well which,
however, are again linked to economic conditions. Undoubtedly, the
migratory movements of an underdeveloped nation are much more strongly
affected by climate change than those of a highly developed nation,
reflecting the rhythmic variations of the climate in the ebb and flow of their
migrating people. 453
453 Compare the following publications:
Briickner, Ed., Klimaschwankungen seit 1700 [Climate change since 1700], Wien, 1890.
[Partly reprinted in Chapter 4 of this anthology]
Uber die praktische Bedeutung der Klimaschwankungen [The practical implications of
climate change], Compte rendu du Vme Congr. inter. des Sc. geogr. Bern, 1892, p. 618 fT.
RufJlands Zukurift als Getreidelieferant [Russia's future as grain exporter], Insert of the
Miinchener Allgemeine Zeitung of Nov 19, 1894.
Der EinflufJ der Klimaschwankungen auf die Ernteertriige und Getreidepreise in Europa
[Influence of Climate Variability on Harvest and Grain Prices in Europe] Geographische
Zeitschrift, 1,1895, p. 39 fT. [reprinted in Chapter 6]
Zur Frage der 35jiihrigen Klimaschwankungen [An Inquiry About the 35-Year-Period
Climatic Variations], Petermann's Mittheilungen, 1902, p. 173 fT. [reprinted in Chapter 8]
Klimaschwankungen und Volkerwanderungen im XIX. Jahrhundert. [Climate variations
and mass migration in the XIX century] Internationale Wochenschrift fUr Wissenschaft,
Kunst und Technik, 1910, p. March.
nico.stehr@zu.de
Chapter 11
The Settlement of the United States as Controlled by
Climate and Climate Oscillations·
In the two months of our Transcontinental Excursion, arranged in so
hospitable a manner by the American Geographical Society, each of the
European members visiting America for the first time has without doubt
obtained impressions of the greatest value and so numerous that a long time
would be necessary to work them out. I think one of the most important
impressions was the observation of the enormous differences in the climatic
conditions of the various parts of the United States. We started from the
humid east, where woods and meadows cover the ground and an abundant
agriculture is possible. A ride of one night brought us from Duluth with its
great woods to the borders of the prairies in the neighborhood of Fargo.
Nevertheless rich crops of wheat are harvested here. As the rainfall
diminishes to the west the grain fields gradually disappear, and in the "bad
lands" of the Little Missouri we were in a half-desert region, where
vegetation requires a regular water supply, in so far as it is found only along
the rivers. Much of the western region is of this character and may be called
half-desert. Regions that are fully desert are rare. We have seen such regions
only in the Great Basin on the bottom of the extinct Lakes Lahontan and
Bonneville. Fortunately only the plains and the lower mountains show these
features. The higher mountains, on the other hand, are able to condense the
vapor of the air and therefore enjoy a greater rainfall.
Nevertheless the half-desert regions today are to some extent inhabited
and now yield good crops. Man has by his skill and by his work in many
cases made out of a desert a paradise. In the neighborhood of Coulee City on
the Columbia Plateau we saw the admirable results of dry farming. By an in-
genious choice of the sequence of grain and by a not less ingenious use of
the natural water supply furnished by rain, crops are produced, not very rich,
but giving a good return to the farmer who practices extensive agriculture.
Presented on October 18, 1912, at the scientific meetings held after the return of the Trans-
continental Excursion to New York.
299
nico.stehr@zu.de
300 EDUARD BRUCKNER
Much greater success attends his efforts where it is possible to utilize, in
the streams that flow down from them, the water fallen as rain in the
mountains, for the irrigation of the arid and semi-arid plains. Where once
was a desert we now see rich orchards. During the Excursion we saw
extensive oases of this nature, which had developed in a few years in the
Yakima valley, Washington, at the Dalles on the Columbia River, at Salt
Lake City, at Grand Junction, Colorado, and at Phoenix, Arizona. Indeed
fruit trees find here the best conditions that can be imagined. In humid
regions the water supply by rainfall decreases the temperature of the air by
diminishing the amount of radiation from the sun because of clouds. In the
irrigated regions of the West that is not so: the fruit trees receive their water
from beneath without an interruption of the radiation of the sun.
Wonderful results are obtained by the co-operation of governmental and
private work, the government studying through its Geological Survey and its
Reclamation Service the available water supply and making it possible of
utilization by building canals, reservoirs, etc. Today the water available for
irrigation is not yet exhausted. In the Yakima valley the government will
provide the water supply for 34,000 acres beyond the area now under
cultivation. In the neighborhood of Phoenix, 160,000 acres are under
irrigation today. But the ar-ea might be increased by the water supply already
available to 230,000 acres, of which 190,000 can be irrigated directly by
surface supply, regulated by the Roosevelt Dam, and 40,000 by pumping
ground water.
Surely today an increase in population by using to a greater extent the
water available for irrigation is possible in some regions, but not indefmitely.
The available water is controlled by climate, and, therefore, there is a limit
beyond which man cannot go. But still more. This limit would be constant
only if rainfall and the other meteorological conditions that control the water
supply, or-in other words--only if climate were constant. Now we have
proofs that climate is not constant but that there are climatic oscillations of
importance that affect to a marked degree rainfall and temperature.
Some years ago I showed that such oscillations of climate are observable
over the whole world. They consist, on the continents, of an alternation of
relatively warm and dry with relatively cool and humid periods. Taking only
the last century, we have series of humid and cold years about 1815, 1850,
and 1885, series of warm and dry years about 1830, 1865, and 1900.
In an extensive paper published in 1890,454 I have worked out the meteo-
rological observations of about 800 stations, distributed over the whole
454 The German titles are as follows: Ed. Bruckner: Klimaschwankungen seit 1700 nebst
Beobachtungen uber die Klimaschwankungen der Diluvialzeit [Climate change since 1700
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 301
world. I proved that climatic oscillations are simultaneous on the whole
earth, oscillations of temperature being the initial cause. The oscillations of
temperature cause oscillations of the distribution of air pressure. During a
warmer period, the pressure is distributed in such a manner that the overflow
of air from the Ocean to the continents diminishes, while during a cooler
period, on the contrary, it increases. Therefore during a cooler period the
continents receive more humid air and more rain than during a warmer
period. Coolness and humidity coincide on the continents, while the oceans
and also some of the coastal regions of the continents receive less rainfall
during the cool periods.
It is not possible to reproduce here the detailed tables, given in the book
mentioned above, that illustrate climatic oscillations up to 1885. I will only
give some examples of the oscillations of rainfall.
They are shown by the curves of Figure 1. They give the rainfall of
various meteorological stations that represent different regions of the world:
Brussels represents western Europe; Bremen central Europe; Nertschinsk,
the central part of eastern Siberia; Nikolaevsk on the Amur, the Pacific coast
of Siberia; and Madras, India. European Russia is represented by the average
of a great number of stations in the basin of the Don River. The United
States is represented by three curves, one calculated from the observations of
a great number of stations in the upper Ohio Valley, the second calculated in
the same manner for the central part of the Mississippi Valley, and the third
for New England. I do not give the data for each year, because there are
many minor irregularities due to local thunder storms, local heavy rains, etc.
But for each year I use the average of the ten years of which it represents the
etc.; Chapter 4 of this anthology], Wi en 1890. 324 pp.; see also the following papers
treating on chmatic oscillations:
Ed. Bruckner: Ober die Bedeutung der Klimaschwankungen for das praktische Leben [The
practical implications of climate change], Compte rendu du Vme Congr. intern. des Sc.
geogr., Bern, 1892, pp. 616--623
- Ruj3lands ZukunJt als Getreidelieferant [Russia's future as grain exporter ], Beilage zur
Miinchener Allgemeinen Zeitung Nov. 19,1894.
- Der Einjluss der Klimaschwankungen auf die Ernteertriige und Getreidepreise in Europa
[Injluence of Climate Variability on Harvest and Grain Prices in Europe], Geographische
Zeitschrift, Vol. 1,1895, pp. 39-51,100-108.
- Zur Frage der 35jiihrigen Klimaschwankungen [An Inquiry About the 35-Year-Period
Climatic Variations; reprinted in Chapter 8], Petermann's Mittheilungen, 1902, pp.
173-178.
- Klimaschwankungen und Viilkerwanderungen im XIX Jahrhundert [Climate variations
and mass migration in the XIX century], Internationale Wochenschrift flir Wissenschaft,
Kunst und Technik, 1901, March 5.
- Klimaschwankungen und Viilkerwanderungen. [Climate variations and mass migration;
reprinted as Chapter 10], Vortrag gehalten in der feierlichen Sitzung der K. Akademie der
Wissenschaften am 31. Mai 1912. Wien, 1912, 24 pp.
nico.stehr@zu.de
302 EDUARD BRUCKNER
center, viz.: for the year 1835, the average of the years 1831-40; for 1836,
the average of 1832-41; for 1837, the average of 1833-42, etc.
1830 40 50 60 70 80 90 1900
I
r-Brlissel
~
- .r
""- "
" V-
-'
/ V"'" '\ / ~
Bremen ,V"\... ./ "~
-
Don_ ~ ~
, "'" ~
Nerchlnsk,
~~
./ '-' ~
r ...
rJ\.
Nikolaevsk
- 'r /\
"
r
Madras
1/ \ ,- ~ \
/ \ .... lA.1
IV ~ V
"\
/ \ / 1\.-'" ~I\.
- \r
Ohio
\/ " v
\,
v
Mississippi,
r
l-
V
~ \
'",.V \.. ~
-
New England ~~ -, ~ /' \
\~
I
\
~
JIf'/
\J 'V
-
1830 40 50 60 70 80 90 1900
Figure 1I .1. Oscillation of Rainfall in Various Regions of the World. Horizontal divisions =
100 mm rainfall
It will readily be seen that in all parts of the world represented in the
diagrams, with the single exception of New England, there is a maximum of
rainfall about 1845-50, a minimum about 1860-70, another maximum about
1880, followed by a decrease of rain until the end of the last century. The
oscillations are rather great, the difference between the maximum and the
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 303
minimum amounting, in the most continental regions of the earth, to 50
percent and more. Only the curve representing the fluctuation of rainfall in
New England has another rhythm, having maxima in 1869 and in 1889; but
we must keep in mind that the climate here is quite oceanic, and therefore
the oscillations of rainfall follow the oscillations of rainfall on the Ocean,
which are the converse of the oscillations on the continents, as stated above.
This raises the question: Are these oscillations of climate felt in the
history of the settlement of the United States?
One of the most characteristic features in the distribution of the
population in the United States is the displacement of the center of
population in a westerly direction. The censuses of the different decades
show this very clearly. It is caused by the rapid settlement of the Far West,
which took place principally in the 70s and 80s of the last century. The
following table shows this. It concerns the states west of the first tier of
states adjoining the Mississippi on the west. The increase in population is
given in percentage of the population at the beginning of each decade. For
example, in Montana the increase of population from 1870 to 1880 was 91 %
of the population in 1870; from 1880 to 1890, 238% of the population in
1880, etc.
Table 11.1. Increase ofPoEulation {%} in the United States from 1870 to 1900
1870-80 1880-90 1890-1900
Montana 91 238 70
Idaho 117 118 83
Wyoming 128 192 48
Nevada 46 -26 -11
Utah 66 44 31
Colorado 388 112 31
Arizona 319 47 39
New Mexico 30 29 22
North Dakota 1435 395 67
South Dakota 734 235 15
Nebraska 267 134 0
Kansas 173 43 3
Texas 94 40 36
North Atlantic Division 18 20 21
South Atlantic Division 30 17 18
North Central Division 34 29 18
South Central Division 39 23 26
Western Division 79 71 32
The table demonstrates very clearly the rapid settlement of the Far West
and the enormous growth of the population from 1870 to 1880 and also from
1880 to 1890. But afterwards the growth became slow.
nico.stehr@zu.de
304 EDUARD BRUCKNER
In the table are also given the corresponding data for the great divisions
of the United States used in the census reports. It is very easy to see that the
enonnous growth from 1870-90 is confined to the western states and that the
retardation of the growth since 1890 is also marked only in the western
states. This phenomenon is without doubt due in large measure to the fact
that the population in the western states had grown by immigration from
1870-90 so much as to fill the region completely or nearly completely. But
one point seems to me to have been overlooked: the rapid settlement of the
Far West from 1870-90 coincided with the period of great humidity. The
rainfall in this period was relatively great; the years in which it exceeded the
average were especially numerous, more so than before. Since 1890 there
has come a dry period. I suppose that the desiccation of the climate since
1890 must also be taken into account if we wish to explain the small increase
of the population since 1890, the drier weather resulting in poor crops.
Again, because of the dry period that began in 1890 the population of the
Far West was obliged to abandon large areas that were settled in the
preceding wet period. In the wet period there was a sufficient water supply
for the fanner-not so at the end of the last and at the beginning of our
century, with its smaller rainfall. The diminution of the inhabited area is very
clearly to be seen on the maps of density of population published in the
United States census reports. I have measured upon the map giving the
density of population for 1890 the area that had a population of at least 2 per
square mile. This measurement I repeated upon the map for 1900. The
following table gives the comparison of the results. The states with an
increase in the inhabited area are designated plus; those with a decrease,
minus. Because the inhabited region is of different size in the different states
I have also calculated the decrease in the inhabited area in percentage of the
total area of each state.
Most of the states show a very conspicuous decrease in the area with a
popUlation of 2 or more per square mile or, in other words, in the inhabited
area. If we take all thirteen states together, we find a total decrease in the
inhabited area of 242,800 square kilometers (l00,000 square miles in round
numbers), or of 15% of the inhabited area of 1890. That is a great deal and is
very significant.
If we seek the reason for this remarkable phenomenon of the ebb of the
population of the exposed parts of the Far West, we must bear in mind that it
was especially the thinly populated regions that suffered the decrease. There
is therefore no doubt that the recession of the population is accompanied by
a decrease in the area used for agricultural purposes.
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 305
Table 11.2. Variation in the inhabited area in the Western States from 1890 to 1900
Inhabited Area Variation 1890-1900
in Sq. Kilometers
1890 S9. km Per Cent
Montana 130,000 -15,000 -12
Idaho 101,900 -59,200 -58
Wyoming 61,300 -9,600 -16
Nevada 32,400 -27,900 -86
Utah 60,400 -2,100 -3
Colorado 199,100 -68,800 -35
Arizona 34,700 -5,800 -17
New Mexico 114,200 -69,800 -61
North Dakota 50,800 +38,500 +76
South Dakota 113,500 +1,100 +1
Nebraska 153,400 -6,100 -4
Kansas 186,000 -22,100 -12
Texas 368,800 +14,200 +4
Total 1,638,700 -242,800 -15
But what is the reason for the retreat of agriculture? Evidently the reason
is somewhat different in the different states. In states like Nevada or a part of
Colorado, where agriculture depends closely upon mining because it is espe-
cially for the use of the mining districts that grain is here cultivated-and
that with great difficulty-the decrease in mining activities has also caused a
decrease in agriculture. But in states where agriculture is of prime
importance, such as Kansas, Wyoming, and Nebraska, I am inclined to think
that the decrease in agriculture is due to the dry period that began in 1890. In
these states, it is reported, one may often see abandoned farms as a sign of
the retreat of agriculture. Nevertheless the population of the states as a whole
has not diminished; it is only concentrated in the more favorable regions and
in the towns, while the drier regions have been abandoned. One exception to
the rule is seen in Dakota. Here no sign of a retreat of the popUlation is
apparent; on the contrary, here the inhabited area has increased from 1890 to
1900. To be sure, in 1890 this region was very far from being fully
populated. In Kansas and Nebraska, on the contrary, wide areas had been put
under the plough in the wet period which, in the following dry period, failed
to yield crops. This is a very clear instance of who climatic oscillations
control settlement near its border toward the desert.
Now we are in the beginning of a new humid period. The rainfall in the
United States has been increasing during the past few years. About 1920 we
may expect a maximum of humidity. For there can be no doubt that the
oscillations of climate will continue, since they have been followed back to
Europe over 700 years, each oscillation having from maximum to maximum
a duration of about 35 years. Therefore it must not be forgotten that the
nico.stehr@zu.de
306 EDUARD BRUCKNER
irrigation works which are now under construction have better conditions
than in the middle of a dry period. In the next dry period the rainfall will be
less; therefore the flow of water and the water available for irrigation will
also be diminished. For irrigation projects of long duration it is necessary to
keep this in mind. They should be adapted to the water supply of dry
periods. Therefore the climatic conditions of the dry periods will control the
areas of permanent settlement. Outside of these there always will be a region
fit for settlement in the wet period, but uninhabitable in the dry. Doubtless,
in many districts that are now being irrigated, the limit imposed by the water
supply of the dry period has not yet been reached. In other districts it may be
already. Here the next dry period will first show that the settlements are
controlled in a very high degree by the oscillations of climate.
In the case of Great Salt Lake, the oscillations of climate might influence
human activity in other respects. The level of Great Salt Lake is not constant.
As the lake has no outlet and the water brought in by streams and by rainfall
is entirely absorbed by evaporation, it is highly dependent upon climatic os-
cillations. Since the close of the dry period just after the middle of the last
century, the lake rose more than 12 feet up to about 1880, with a maximum
during the wet period. When the dry period came, it fell to a low level. But
now it is rising again. During our excursion, Mr. Hood, Chief Engineer of
the Southern Pacific Company, gave us a table showing that the level of the
lake was low at the beginning of the present century, but that since 1905 or
1906 it has risen no less than 6 feet. There can be no doubt that these
oscillations of the lake are due to the oscillations of climate. I am aware that
in general some other causes have Now the Union Pacific Railroad has built
an embankment across the lake to shorten the road. The top of this
embankment with the rails is today only a few feet above the level of the
lake; I estimated from the window of our train about 5 feet, no more. If the
future rise of the lake from the dry to the wet period will be as great as it has
been before the maximum of about 1880, we must expect that in the middle
of the next wet period, which we may expect about 1920, the level of the
lake will be 6 feet higher than today and therefore will just cover the tracks.
But in another way also the settlement of the United States is controlled
by climatic oscillations. At the beginning of this paper I showed by a series
of curves (Fig. I) that the oscillations of rainfall in Europe and in the United
States coincide and that the two continents have simultaneous wet and dry
periods. Now, the weather influences crops, but in a very different manner.
In the United States the wet years are the good years, because in the cereal
region with its high summer temperature, the rainfall is in general not very
abundant. On the contrary, in the western part of Europe, including
Germany, the dry years are good years since here in the cool climate the
crops suffer because of too much humidity. The following Figures 2 and 3
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 307
may elucidate the relation between rainfall and crops in the western part of
Europe. Above and beneath the diagrams are given the years, each year
representing the middle of a lustrum, viz. 1848 the lustrum 1846-50, 1853,
the lustrum 1851-55, etc. The curves marked R give the fluctuation of
rainfall in percentage of the average. In Figure 2, WZ means crop of wheat
expressed in percentage of an average crop; RO means crop of rye.
1848 53 58 63 68 73 78 83 88 91
~ "-.
"I"-
4%
R /
"- "",- /'
/ 0%
--"
,/ -4%
wz r---
o
RO
-, , r--.
"- L
i"- " ./ o
" -
"'~
",.,."".
1848 53 58 63 68 73 78 83 88 91
Figure 11.3: Relation of Rainfall (R) to Crops (WZ and RO) in Prussia. Grain
production and rainfall are indicated in percentages of the variation from their respective
means, which are based on a long series of observations. One division of the former is
equivalent to 5% variation; of the latter to 4%.
1758 63 68 73 78 83 88 83 98180308 13 18 23 28 33 38 43 48
34M
30M 17Fr
26~f 15Fr
+41 I3Fr
+ 2t Oil
Ot
- 2t lOCk
1758 63 68 73 78 83 88 83 981803 08 13 18 23 28 33 38 43 48
Figure 11.4 . Relation of Rainfall (R and E) to Price of Grain (WM and WZ) in Central
Europe. WM = price, in marks, of a Bavarian bushel of wheat in Munchen (I division = 2
marks); WZ = price, in francs, of 50 kilograms of wheat in Zurich; R = rainfall in South
Germany (1 division = 3% variation from the mean of many years); E = date of the vintage in
nico.stehr@zu.de
308 EDUARD BRUCKNER
South Germany and Switzerland, which varies proportionally with the rainfall, being late in
moist and early in dry years (I division = 1 day; the mean of many years is marked 0 t).
Figure 2 shows that, when the rainfall increases in Prussia, the crops
decrease and vice versa. The crops were good in the dry years about 1860,
but very bad in the wet years about 1880; in the following drier period they
increased again.
As the crop statistics do not go back very far, I have taken the price of
grain as a criterion for the quality of the crop; for, before an international
commerce in grain existed, the price of grain in a country was always a
function of the crop of that country. When the crop was bad, the price was
high and vice versa. In Figure 3 the curve WM shows the variation of the
price of wheat at Miinchen (marks per Bushel); WZ, of the price of wheat at
Ziirich (francs per Bushel). E means the date of the vintage, which is a good
measure of the rainfall, the grapes being harvested earlier in the dry years, as
I have shown in the paper mentioned above.
The figure shows at a glance that in the wet periods about 1770, 1810,
and 1845 grain was dear; in the dry periods about 1785 and 1830, cheap.
In the continental climates of Europe the relation between rainfall and
crops is the same as in the United States. In Russia the crop of rye increases
and decreases with the rainfall as in Ohio (see Fig. 4).
1853 58 63 68 73 78 83 88
~
RO
~ '\. Ohl
C':S
'Cil
" "-"- , f
f "-
""
'"
~I%
./ """ \..
R
.,./
~
"'
,
..".
"'\. J
'\ L 1 16 Bush
I"- f
W7 J
/
o
~ "" "-
R "'-
./
II"".
f
/
t..,.......-
'""
1853 58 63 68
.JI"
1/
73 78 83
I
I
88
0%
Figure 11.5. Relation of Rainfall (R) to Crops (RO and WZ) in Russia and in Ohio. One
division is equivalent to 2% for the rainfall (R); to 1,500,000 hectoliters for the increase in the
export of rye from Russia (RO); and to 0.8 bushel per Acre of the wheat crop in Ohio (WZ).
(The variation in the increase of the rye export from Russia was determined as follows: A
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 309
straight line was drawn, according to the method of least squares, through the recorded
quantities exported from 1851 to 1890, and the differences for every five years between these
values and those actually observed were then plotted.)
These conditions are of great importance for the emigration from Europe
and the immigration into the United States. The accompanying figure (Fig.
5) gives a diagram showing the oscillation of rainfall in the United States
and in western Europe, and under it a diagram showing the immigration into
the United States, all curves being constructed by using five-year averages.
In general the immigration into the United States increased from the
)beginning to the end of the curve. But the increase was not regular. In the
wet period about 1850, very well-pronounced in western Europe and in the
United States, the immigration into the United States increased; in the
following dry period it decreased. Five years after the 1880 maximum of
rainfall in western Europe the immigration also had a maximum; it decreased
materially in the next dry period. Only after 1900 it increased in an
extraordinary manner.
1833 43 53 63 73 83 93 1903
%
+10 T
Rain ) 1\ United States V~
"\. \,
+5
o
I V
-5 / /
I "-
""
j ~
"\
..,i---""
" " I
+5
o
I
......
I j
350
-5
I- -W-Europc
Rain" / ,~
300
........ J\ I
I ...... I
- 250
.-- I '\
200
150
/ \'
--
j .......
........ 100
I 50
~ ~ Total Immigration into the United States
80
60
II
j
A
\ , V
~
1\ I
~
~
'\
"
40
20 1/ f' V
10' Rri~ish !m~igr~tion, int~ the, Uni,tcd ~tat~s
1833 43 53 63 73 83 93 1903
Figure 11.6. The Fluctuation of Rainfall in the United States and Western Europe as
Compared, Respectively, with the Total Immigration and British Immigration to the United
States. The curves are based on five-year totals and are not adjusted. The number of
nico.stehr@zu.de
310 EDUARD BRUCKNER
immigrants is given in ten thousands, rainfall in deviations (percentages) from the mean of
many years.
The parallelism between rainfall and the emigration from Europe is to be
seen much better when we compare only the curve of rainfall for western
Europe with British emigration (Fig. 5), or the curve of rainfall for Germany
and the curve of emigration from Germany (Fig. 6). The parallelism of these
two sets of curves is striking. The reason for the parallelism is clear: the
greater part of the emigrants coming from Europe to the United States are
agricultural. The rainy period causes bad crops in western Europe, including
Germany, and therefore gives an impulse to emigration. In the same period
the greater humidity is associated with good crops in the United States. This
fact is communicated by correspondence to the relatives of the immigrants
remaining in Europe, and this furnishes an additional incentive to
immigration. As greater rainfall and the bad crops resulting therefrom are the
cause of the increase of emigration out of Europe, it is not astonishing that
sometimes the fluctuations of the curve of emigration are five years behind
the fluctuations of rainfall.
- ...
1833 38 43 48 53 58 63 68 73 78 83 88 93 98 1903
i i f i i
+10% ~Rainfall
+5%
"'-.
0% I/' "- I
...... I"""
-5%
100 " "- ---,. f "-II
80
-" -
60
.........
f "-
40 ./ "- f I "''--
20 ~ ........ ""-
o
-Emigration
I I I
1833 38 43 48 53 58 63 68 73 78 83 88 93 98 1903
Figure 11.7. Fluctuation of Rainfall in Germany and of German Emigration to the
United States. The curves are based on five-year totals and are not smoothed. Rainfall is
indicated in percentage of variation from the mean of many years; the number of immigrants
into the United States, in ten thousands.
Since 1900, the coincidence between rainfall and immigration into the
United States ceases; since then immigration has reached numbers never
attained before. That seems to be an exception to our rule. But if we examine
the composition of this immigration, we find that the immigration from
Great Britain and Germany is now very small, but that Russia and the
eastern parts of Austria and Hungary are sending enormous numbers of
emigrants. Here, where the summer is hot and the rainfall small, as in the
United States, wet years are good years, dry years are bad years. Therefore
nico.stehr@zu.de
THE SETTLEMENT OF THE UNITED STATES 311
in the dry period around 1900 the impulse to immigration has been great. To
be sure, especially in Russia, there are also political considerations to be kept
in mind. They might be of more importance in this connection than bad
crops.
While we have seen that there is a correlation between the oscillations of
climate and immigration into the United States, I am nevertheless far from
overlooking other causes of emigration from Europe. The great density of
population in Europe and the extended room available for settlement in the
United States are constantly at work to induce emigration from Europe, and
political causes are not lacking. But these forces, which are constantly at
work, cannot veil the influence of climatic oscillations. The stream of
immigrants to the United States ebbs and flows with the oscillations of
climate, which give it a rhythmical impulse. And not only is immigration to
the United States controlled by climatic oscillations, but also the settlement
of the Far West, as we have seen.
nico.stehr@zu.de
List of Publications of Eduard Briickner
CLIMATE
1. Uber die Methode der Zahlung der Regentage und deren EinjlufJ auf die
resultierende Periode der Regenhaufigkeit. Meteorologische Zeitschrift,
Volume 3, 1887.
2. Notre climat subit-i/ des changements? Archives des sciences physiques
et naturelles Sept.-Okt., 1888.
3. Die meteorologische Station auf dem Santis. Meteorologische Zeit-
schrift, Volume 5, 1888.
4. In wie weit ist das heutige Klima konstant?455 Verhandlungen des VIII.
Deutschen Geographentages, 1889.
5 . Klimaschwankungen seit 1700 nebst Bemerkungen iiber die
Klimaschwankungen der Diluvialzeit. 456 Pencks Geographische
Abhandlungen, Volume 4, 1890.
6. Verdunstung einer Schneedecke. Meteorologische Zeitschrift, Volume 7,
1890.
7. Das Klima der Eiszeit. Verhandlungen der 73. Jahresversammlung der
Schweizer Naturforschenden Gesellschaft, Davos, 1891.
8. Uber die Bedeutung der Klimaschwankungen for das praktische Leben.
Compte rendu du V. Congres International des Science Geographie,
1891.
455 Chapter 3 in this book
456 Excerpts in Chapter 4
313
nico.stehr@zu.de
314 EDUARD BROCKNER
9. Materialien zur Verfolgung mehrjiihriger oder siikuliirer Perioden der
Witterung. Meteorologische Zeitschrift, Volume 9,1892.
10. Uber den Einfluft der Schneedecke aUf das Klima der Alpen. 457 Zeit-
schrift des Deutschen und Osterreichischen Alpen-Vereins, 1893.
11. Diirren in Ostasien. Meteorologische Zeitschrift, Volume 11, 1894.
12. Das Klima von Odessa. Meteorologische Zeitschrift, Volume 11, 1894.
13. Meteorologische Stationen in den Franzosischen Alpen. Meteorolo-
gische Zeitschrift, Volume 12, 1895.
14. Der Einfluft der Klimaschwankungen auf die Ernteertriige und Getrei-
depreise in Europa. 4S8 Geographische Zeitschrift, Volume 1, 1895.
15. Uber die Herkunft des Regens. Verhandlungen des VII. Internationalen
Geographenkongresses in Berlin 1899.
16. Uber die Herkunft des Regens. Geographische Zeitschrift, Volume 6,
1900.
17. Zur Frage der 35-jiihrigen Klimaschwankungen. 459 Petermann's Mitt-
heilungen, 1902.
18. Wetterpropheten. ~60 20. Jahresbericht der Geographischen Gesellschaft
Bern, 1903/04.
19. Hohengrenzen in der Schweiz, Naturwissenschaftliche Wochenschrift,
N. F., Volume 4 (Volume 20) No. 52, 1905.
20. Die Bilanz des Kreislaufs des Wassers auf der Erde. Geographische
Zeitschrift, Volume 11, 1905. (In the same year also published in
Russian.)
21. Meer und Regen. Naturwissenschaftliche Wochenschrift, N.F., Volume
4, No. 26,1905.
22. Schwankungen des Niederschlages im Deutschen Reich 1816-1900.
Meteorologische Zeitschrift, Volume 23, 1906.
23. Schwankungen des Niederschlages im Deutschen Reich. Zeitschrift flir
Gletscherkunde, Volume 1, 1906/07.
24. Niederschlag, AbjlujJ und Verdunstung auf den Landjliichen der Erde.
Meteorologische Zeitschrift, Volume 25, 1908.
25. Schnee in der algerischen Sahara. Zeitschrift flir Gletscherkunde,
Volume 11, 1908.
26. Uber Klimaschwanklingen. 461 Mitteilungen der Deutschen Landwirt-
schaftsgesellschaft, Volume 24, 1909.
27. Klimaschwankungen und Volkerwanderungen im XIX Jahrhundert. In-
ternationale Wochenschrift fUr Wissenschaft, Kunst und Technik, 1910.
457 Chapter 5
458 Chapter 6 in this collection
459 Chapter 8
460 Chapter 7
461 Chapter 9 in this collection
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 315
28. Prozentischer Anted des Schnees am gesamten Niederschlag in
verschiedenen Hohen der Schweiz. Zeitschrift fur Gletscherkunde,
Volume 5,1910/11.
29. Uber die Klimaschwankungen der Quartiirzeit und ihre Ursachen.
Compte rendu XI. Congn!s geol. international Stockholm 1910 (1912).
30. Ergebnisse der Schneemessungen in den Schweizer Hochalpen. Zeit-
schrift fur Gletscherkunde, Volume 5, 1910/11.
31. Klimaschwankungen und Volkerwanderungen. 462 Vortrag in der feier-
lichen Sitzung der kaiserlichen Akademie der Wissenschaften, Wien
1912.
32. Moorbildungen und postglaziale Klimaschwankungen am Nordsaum der
Ostalpen. Zeitschrift fur Gletscherkunde, Volume 7, 1912113.
33. With Albrecht Penck: Uber die Verschiebung der Klimagurtel in der
Quartiirzeit. Zeitschrift fur Gletscherkunde, Volume 9, 1914/15.
34. The settlements of the U.S. as controlled by climate and climatic
oscillations. 463 Memorial-Volumen of the transcontinental excursion of
1912 of the American Geographical Society of New York 1915.
35. Klimaschwankungen 1813-1912 in Vorderindien. Festschrift "Albrecht
Penck",1918.
36. /fnderungen der geographischen Breiten und des Klimas in geologischer
Zeit. Zeitschrift fur Gletscherkunde, Volume 12, 1921122.
37. Verdunstung und Kondensation an Schnee und Eis im Gebirge.
Zeitschriftfor Gletscherkunde, Volume 12, 1921122.
38. Die Schneedecke an der StraJ3e zum GroJ3en St. Bernhard in Wallis.
Zeitschrift fur Gletscherkunde, Volume 12, 1921122.
39. With W. Koppen und A. Wegener: Die Klimate der geologischen
Vorzeit. Zeitschrift fur Gletscherkunde, Volume 14, 1925/26.
40. Forschungen uber das diluviale Klima in Mexiko. Zeitschrift fUr
Gletscherkunde, Volume 15, 1926127.
GLACIERS
1. Die Hohen Tauern und ihre Eisbedeckung. Eine orometrische Studie.
Zeitschrift des Deutschen und Osterreichischen Alpen-Vereins 1886.
2. Die Hohe der Schneelinie und ihre Bestimmung. Meteorologische
Zeitschrift, Volume 4, 1887,
3. E. Richters Untersuchungen uber die Schwankungen der Alpengletscher.
Volume 1891.
462 Chapter 10
463 Chapter 11
nico.stehr@zu.de
316 EDUARD BRUCKNER
4. Der Gletscherabbruch an der Altels im Berner Oberland vom 11. Sept.
1895. Himmel und Erde, Volume 8, 1896.
5. Mit S. Finsterwalder: Protokoll der 3. Internat. GletscherkonJerenz in
Maloja 6. bis 9. Sept. 1905. P. M., Volume 51,1905.
6. Die Hohe der Firnlinie am Hujigletscher und die Methode der Bestim-
mung der Hohe der Firnlinie im allgemeinen. Vierteljahrsschrift der
Naturforschenden Gesellschaft Zurich 1906, Volume 51.
7. Zur Einfiihrung. Zeitschriftfiir Gletscherkunde, Volume 1, 1906/07.
8. Mit E. Muret: Les variations periodiques des glaciers. XII. Rapport,
1906. Zeitschrift fUr Gletscherkunde, Volume 2, 1907/08
9. Die SchmelzJormen des Firns im tropischen und subtropischen
Hochgebirge. (Nieve Penitente). Zeitschrift fUr Gletscherkunde, Volume
2, 1908.
10. Schneeschmelzkegel in den Alpen und Nieve Penitente. Zeitschrift fUr
Gletscherkunde, Volume 2, 1908.
11. Mit E. Muret: Les variations periodiques des glaciers. XIII. Rapport,
1907. Zeitschrift fUr Gletscherkunde, Volume 3, 1908/09.
12. Mit E. Muret: Les variations periodiques des glaciers. XIV. Rapport,
1908. Zeitschrift fUr Gletscherkunde, Volume 4, 1909/10.
13. Gletscherschwankungen in der Schweiz 1800-1900. Zeitschrift fUr Glet-
scherkunde, Volume 4, 1909/10.
14. GroJ3e der Ablation am Rhonegletscher. Zeitschrift fUr Gletscherkunde,
Volume 5, 1910/11.
15. Bericht der internationalen Gletscherkommission fur die Jahre
1907-1910. Zeitschrift fUr Gletscherkunde, Volume 5, 1910/11.
16. Mit E. Muret: Les variations periodiques des glaciers. XV. Rapport,
1909. Zeitschrift fUr Gletscherkunde, Volume 5, 1910/11.
17. Die Firnhaube des TitlisgipJels in den Glarner Alpen, Schweiz.
Zeitschrift fUr Gletscherkunde, Volume 5, 1910/11.
18. Bemerkungen zu der Abhandlung des Herrn Lamansky uber das
Absterben der Gletscher. Zeitschrift fUr Gletscherkunde, Volume 8,
1913/14.
19. Bolletino del Comitato Glaciologico Italiano. Zeitschrift fUr
Gletscherkunde, Volume 9, 1914/15.
20. Beobachtungen uber die GroJ3enanderungen der Gletscher der Ostalpen
in den Kriegsjahren 1914 und 1915. Zeitschrift fur Gletscherkunde,
Volume 9, 1914/15.
21. Veranderungen im Stande der Gletscher der osterreichischen Alpen
nach den Beobachtungen der Jahre 1914, 1915, und 1916. Zeitschrift
fUr Gletscherkunde, Volume 10, 1916/17.
22. Eigenartige stengelige Anordnung des Eises in einem Glaszylinder.
Zeitschrift fUr Gletscherkunde, Volume 10, 1916/17.
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 317
23. Das Vorrucken der Gletscher in den Ostalpen. Mitteilungen des
Deutschen und Osterreichischen Apenvereins, 1917.
24. Die Forderung der Wissenschaft von den Alpen durch den deutschen
und osterreichischen Alpenverein in den letzten 25 Jahren. Zeitschrift
des Deutschen und Osterreichischen Alpen-Vereins 1919.
25. With Sikosak,: 1m Meer gronlandischer Fjorde entstandene Firnfelder
und Gletscher. Zeitschrift fur Gletscherkunde, Volume 11, 1918/20.
26. With J. Bowman: Uber Schneerosion und Entstehung der Kare.
Zeitschrift fur Gletscherkunde, Volume 12, 1921 .
.27. VorstoJ3 der Schweizer Gletscher. Zeitschrift fUr Gletscherkunde,
Volume 12, 1921.
28. Die meteorologischen Ursachen des GletschervorstoJ3es in den
Schweizer Alpen. Zeitschrift fur Gletscherkunde, Volume 12, 1921.
29. Vermessung des Lysgletschers auf der Sudseite des Monte Rosa.
Zeitschrift fur Gletscherkunde, Volume l3, 1923/24.
30. Die Vergletscherung der Kette Peters des GroJ3en im Pam irgeb iet.
Zeitschrift fur Gletscherkunde, Volume l3, 1923/24.
31. Fortbewegung des "Hotel des Neuchiitelois" auf dem Unteraargletscher
im Berner Oberland 1842-1922. Zeitschrift fur Gletscherkunde, Volume
14,1925/26.
32. Verzeichnis der rezenten Gletscher Italiens. Zeitschrift fur Gletscher-
kunde, Volume 15, 1926/27.
33. Dickenmessungen von Gletschern mittels seismischer Methoden.
Zeitschrift fur Gletscherkunde, Volume 15, 1926/27.
34. Mit o. Liitschg: Uber Gletscher, Niederschlag und AbjluJ3 im Mattmark-
gebiet (Monte Rosagruppe). Zeitschrift fur Gletscherkunde, Volume 15,
1926/27.
GLACIAL AGES
1. Die Eiszeit in den Deutschen Alpen, nach E. Penck. Kosmos, I. Volume,
1884.
2. Uber die Vergletscherung Ostsibiriens. N. Jahrbuch fur Mineralogie,
etc., Volume I, 1885.
3. Die Eiszeit am Nordabhang der Alpen. Der Naturforscher, Volume 19,
1886.
4. Die Vergletscherung des Salzachgebietes nebst Beobachtungen uber die
Eiszeit in der Schweiz. Pencks Geographische Abhandlungen, Volume 1,
H. I, 1886.
5. Die Eiszeit in den Alpen. Mitteilungen Geographischen Gesellschaft
Hamburg, 1887/88.
nico.stehr@zu.de
318 EDUARD BRUCKNER
6. Eiszeit-Studien in den sudostlichen Alpen. 10. lahresbericht der Geo-
graphischen Gesellschaft Bern, 1891.
7. Le Systeme glaciaire des Alpes. Guide publie a l'occassion du Congres
Geologique International 6me Session a Zurich en 1894. Bull. Soc. des
scienc. natur. de NeucMtel, T. 22, 1893/94.
8. Die Eiszeiten in den Alpen. Verhandlungen Gesellschaft Deutscher
Naturforscher und A.rzte, 1904. AUg. Teil.
9. Die Eiszeiten in den Alpen. Geographische Zeitschrift, Volume 10, 1904.
10. Die Eiszeiten in den Alpen und die "Einheitlichkeit" der Eiszeit. Geogra-
phische Zeitschrift, Volume 11, 1905.
11. Alpen und Eiszeit. Mitteilungen der Geographischen Gesellschaft
Hamburg, Volume 21, 1906
12. Mit A. Penck: Die Alpen im Eiszeitalter. 2. Volume, 2. Buch, 2. Kap.:
Linth- Reuj3-, Aare- und Rhonegletscher auf Schweizer Boden, 1903/04;
3. Volume, 3. Buch: Eiszeit in den Sudalpen, Brenta-, Piave- und
Isonzogletscher, 1908; 4. Buch: Savegletscher. Work completed in 1909.
13. Palaeozoische Eiszeitspuren in der Kapkolonie. Zeitschrift fUr Glet-
scherkunde, Volume 5, 1911.
14. Uber die Klimaschwankungen der Quartiirzeit und ihre Ursachen.
Compte rendu XI. Congres geol. international Stockholm 1910 (1912).
15. Interglaziale Torjlager in den nordlichen Ostalpen. Zeitschrift fUr
Gletscherkunde., Volume 7, 1912/13.
16. Zur Frage der Verschiebung der Eisscheide in Skandinavien. Zeitschrift
fUr Gletscherkunde, Volume 8,1913/14.
17. Lagerungsverhiiltnisse und Alter der Hottinger Breccie bei Innsbruck.
Zeitschrift fUr Gletscherkunde, Volume 10, 1916/17.
18. Bemerkungen zum Aufsatz von Prof Deecke uber die tiefliegenden glazi-
alen Reste in Sudwestdeutschland und uber die Loj3tratigraphie
Suddeutschlands. Zeitschrift fUr Gletscherkunde, Volume 11, 1918/20.
19. With G. v. Zahn: Uber die angeblichen Moriinen im Thuringer Waldo
Zeitschrift fUr Gletscherkunde, Volume 12, 1921122.
20. Geochronologische Untersuchungen uber die Dauer der Postglazialzeit
in Schweden, in Finnland und in Nordamerika. Zeitschrift fUr Gletscher-
kunde, Volume 12, 1921122.
21. Die glaziale Entwicklungsgeschichte Nordwestskandinaviens. Zeitschrift
fur Gletscherkunde, Volume 12, 1921122, 13, 1923/24.
22. Albrecht Pencks neue Untersuchungen uber die Eiszeit in den nord-
lichen Alpen. Zeitschrift fur G1etscherkunde, Volume 13, 1923/24.
23. Die Salpausselkii-Randbi/dungen des In/andeises in Finn/and. Zeit-
schrift fUr G1etscherkunde, Volume 13, 1923/24.
24. Ein Institut for EiszeitJorschung. Zeitschrift fur G1etscherkunde, Volume
13, 1923/24.
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 319
25. Geochronologische, geomorphologische und pjlanzengeographische
Untersuchungen im Bereich des alten Ragundasees in Jiimtland,
Schweden. Zeitschrift fur Gletscherkunde, Volume 14, 1925/26.
26. Geochronologische Untersuchungen uber den Ruckzug der letzten
Vergletscherung in den Neu-Englandstaaten. Zeitschrift fur Gletscher-
kunde, Volume 14, 1925/26.
27. Die iiuj3erste Grenze der letzten Vergletscherung in Nordwestdeutsch-
land nach Gipp. Zeitschrift fur Gletscherkunde, Volume 14, 1925/26.
28. With H. F. Osborn: Ober die Gliederung des Quartiirs und ihre
Beziehung zur Priihistorie. Zeitschrift fur Gletscherkunde, Volume 14,
1925/26.
29. Die groj3en Endmoriinenzuge Norddeutschlands. Zeitschrift fur Glet-
scherkunde, Volume 15, 1926/27.
30. With Otto Ampferer: Ober geologische Methoden zur Erforschung des
Eiszeitalters. Zeitschrift fur Gletscherkunde, Volume 15, 1926/27.
31. Wieviel Jahre sind seit dem Hochstand der letzten Vergletscherung
verstrichen? Zeitschrift fur Gletscherkunde, Volume 15, 1926/27.
32. "Die Eiszeit". ZeitschriJt for allgemeine EiszeitJorschung usw. Zeitschrift
fur Gletscherkunde. Volume 15, 1926/27.
33. Die geologische und archiiologische Stellung des Hochgebirgspaliioli-
thikums in der Schweiz. Zeitschrift fur Gletscherkunde, Volume 15,
1926/27.
34. With G. H. Steinmann: Ober das Diluvium des Niederrheins und die
Gliederung des Eiszeitalters. Zeitschrift fur Gletscherkunde, Volume 15,
1926/27.
35. Die Ostalpen in der Eiszeit. In Deuticke (Ed): "Die osterreichischen
Alpen". Wien, 1927.
MORPHOLOGY
1. Die feste Erdrinde und ihre Formen. Ein Abrij3 der allgemeinen
Geologie und der Morphologie der Erdoberjliiche. (Volume 2 of
"Allgemeine Erdkunde" von Hann, Hochstetter und Pokorny) g. Aufl.
1897. (1903 a Russian translation by M.A. Engelhardt was published in
Petersburg. )
2. Notice preliminaire sur la morphologie du Jura Suisse et Fram;ais.
Archives des Sciences physiques et naturelles. T. 14, 1902.
3. Morphologie du Plateau Suisse et du Jura. Compte rendu des travaux de
la Soc. helvetique de sc. nat. 1902.
4. Die glazialen Zuge im Antlitz der Alpen. Mitteilungen des Vereins fur
Erdkunde Leipzig 1906 (1907).
nico.stehr@zu.de
320 EDUARD BRUCKNER
5. With W. M. Davis: Uber die glazialen Skulpturformen in Gebirgen.
Zeitschrift flit Gletscherkunde, Volume 2, 1907.
6. With W. Kilian: Uber Glazialerosion und Ubertiefung. Zeitschrift flir
Gletscherkunde, Volume 2,1907.
7. Das Alter der alpinen Landschaftsformen. Jahresbericht der Geographi-
schen Gesellschaft Bern, Volume 21, 1907.
8. Glazialmorphologische Exkursion in das Chamounixgebiet, ins Wallif
und ins Berner Oberland. Compte rendu des travaux du IX. Congres
international de Geographie Geneve 1908, T. 1 (1910).
9. Die glazialen Ziige im Antlitz der Alpen. Naturwissenschaftliche
Wochenschrift N.F., Volume 8, No. 50, 1909 und Compte rendu des
travaux du IX. Congres international de Geographie, Geneve 1908, T. 2
(1910).
10. Zur Frage der Entwicklung der Rhein-Rhone- Wasserscheide. Zeitschrift
der Gesellschaft flit Erdkunde, Berlin 1909.
11. Entstehung der quartiiren Schotterterrassen im Umkreis der Alpen.
Zeitschrift flit Gletscherkunde, Volume 4, 1909/10.
12. Bemerkungen zu V. Hi/ber: Entstehung der quartiiren Schotterterrassen
im Umkreis der Alpen. Zeitschrift flir Gletscherkunde, Volume 4,
1909/10.
13. Gliederung der diluvialen Schotter in der Umgebung von Basel. Zeit-
schrift flir Gletscherkunde, Volume 6, 1911112.
14. Das Zungenbecken des alten Ennsgletschers als Felsbecken. Zeitschrift
fUr Gletscherkunde, Volume 7, 1912/13.
15. Die sogenannten Glazialerscheinungen in der Rhon. Zeitschrift flir Glet-
scherkunde, Volume 8, 1913/14.
16. Zur Frage der Entstehung der Sol/e. Zeitschrift flir Gletscherkunde,
Volume 9, 1914/15.
17. Zur Morphologie der Otscherlandschaft. Mitteilungen der Geographi-
schen Gesellschaft Wien 65,1922.
18. Alte Ziige im Landschaftsbild der Ostalpen. Vortrag Berlin. Zeitschrift
der Gesellschaft flit Erdkunde Berlin, 1922.
19. Zur Glazialmorphologie von Norwegen. Zeitschrift flir Gletscherkunde,
Volume 12, 1921122.
20. Uber die bodengestaltende Wirkung des vorstoftenden Oberen
Grindelwaldgletschers. Zeitschrift flir Gletscherkunde, Volume 12,
1921122.
21. Bemerkungen zu Worm: Kare und Schneegrenze. Zeitschrift flir Glet-
scherkunde, Volume 14, 1925/26.
22. Glazialmorphologische Forschungen in der Schweiz und in Spitzbergen
mittels des Flugzeugs. Zeitschrift fur Gletscherkunde, Volume 14,
1925/26.
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 321
23. Die norwegische Strandplattform und ihre Beziehungen zur Quartarzeit.
Zeitschrift fUr Gletscherkunde, Volume 15, 1926/27.
HYDROLOGY
1. Grundwasser und Typhus. 464 Mitteilungen der Geographischen Gesell-
schaft Hamburg, 1887/88.
2. Ober Schwankungen der Seen und Meere. Verhandlungen des IX.
Deutschen Geographentages, Wien, 1891.
3. Sakulare Schwankungen der Eisverhaltnisse des Hudsonjlusses. Meteo-
rologische Zeitschrift, Volume 9, 1892.
4. Untersuchungen uber die tagliche Periode der Wasserfohrung und die
Bewegung der Hochjluten in der oberen Rhone. Petermann's Mitt-
heilungen, 1895.
5. Bericht der Fluftkommission der Schweizerischen Naturforschenden
Gesellschaft for das Jahr 1896/97. Verhandlungen der Schweizerischen
Naturforschenden Gesellschaft 1897.
6. Bericht der Fluftkommissionfor 1900/1901. Verhandlungen der Schwei-
zerischen N aturforschenden Gesellschaft, 1901.
7. Berichte der Fluftkommission der Schweizerischen Naturforschenden
Gesellschaft pro 1903/04 und 1904/05. Verhandlungen der Schweizeri-
schen Naturforschenden Gesellschaft 1904 und 1905. (Published 1905
and 1906.)
8. Die Bilanz des Kreislaufs des Wassers auf der Erde. Geographische
Zeitschrift, Volume 11, 1905.
9. Bericht der Flufikommissionfor 1905/06. Verhandlungen der Schweize-
rischen Naturforschenden Gesellschaft 1906.
10. Gutachten betreffend die Fo/gen, die die Ausfohrung der Millstatter
Kraftanlage for den See voraussichtlich haben wird. Wien 1908, Ed.
Sieger.
11. Zur Thermik der Alpenseen und einiger Seen Nord-Europas. Geogra-
phische Zeitschrift, Volume 15, 1909.
12. Ober die Temperaturverhaltnisse der Fluftseen, insbesondere der Alpen.
Compte rendu des travaux du IX. Congres international de Geographie
1908, T. 2, (1910).
13. Ober das thermische Regime der Seen mit Abjluft. Verhandlungen der K.
Russ. Geographischen Gesellschaft 47, 1911 (Russisch).
14. Die groften Seen Nordamerikas und ihre Bedeutung for den Verkehr.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 61, 1918.
464 Chapter 1 in this collection
nico.stehr@zu.de
322 EDUARD BRUCKNER
15. Bemerkungen zum Aufsatz von K. Fischer: Die Grundgleichungen des
Wasserhaushaltes der Fluj3gebiete. Meteorologische Zeitschrift, Volume
38, 1921.
OCEANOGRAPHY
1. Die Schwankungen des Wasserstandes im Schwarzen Meere und ihre
Ursachen. Meteorologische Zeitschrift, Volume 2, 1886.
2. Meeresspiegel und Klima. Naturforscher, Volume 20, Tiibingen 1887.
3. Die Schwankungen des Wasserstandes im Kaspischen Meer, dem
Schwarzen Meer und der Ostsee in ihrer Beziehung zur Witterung. 46S
Ann. der Hydrographie und maritime Meteorologie, 1888.
4. Ober Schwankungen der Seen und Meere. Verhandlungen des IX.
Deutschen Geographentages, Wien 1891.
5. Meer und Regen. Naturwissenschaftliche Wochenschrift, N.F, Volume
4, No. 26, 1905.
6. Das italienisch-osterreichische Projekt einer gemeinsamen Erforschung
des Adriatischen Meeres. Mitteilungen der Geographischen Gesellschaft
Wien, Volume 53, 1910.
7. Die erste Kreuzungsfahrt S. M S. "Najade" in der Hochsee der Adria
25. Febr. bis 7. Miirz 1911. Mitteilungen der Geographischen Gesell-
schaft Wien, Volume 54, 1911.
8. Der Zustand des Adriatischen Meeres am Ausgang des Winters 1910111.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 54, 1911.
9. Die dritte Terminfahrt S. M S. "Najade" in der Hochsee der Adria vom
16. Aug. bis 5. Sept. 1911. Mitteilungen der Geographischen Gesell-
schaft Wien, Volume 55, 1912.
10. Das Projekt einer internationalen Erforschung des Mittelmeers.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 57, 1914.
POLAR RESEARCH
1 . Resultate der meteorologischen Beobachtungen der deutschen
Polarstationen 1882183. Meteorologische Zeitschrift, Volume 4,1888.
2. Der Kampf um den Nordpol. Mitteilungen der Geographischen Gesell-
schaft Wien, Volume 52,1909.
3. Filchners deutsche antarktische Expedition. Zeitschrift flir Gletscher-
kunde, Volume 5, 1910/11.
46S Chapter 2 in this collection
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 323
4. Das Eis der Antarktis. Schriften des Vereins zur Verbreitund naturwis-
senschaftlicher Kenntnisse. Volume 51, 1910111.
5. Die Alaskaexpedition der Amerikanischen Nationalen Geographischen
Gesellschaft vom Sommer 1910. Zeitschrift flir Gletscherkunde, Volume
6, 1911.
6. Neuer Plan einer Durchquerung Gronlands. Zeitschrift flir Gletscher-
kunde, Volume 6, 1911/12.
7. Die Schneegrenze in der Antarktis. Zeitschrift flir Gletscherkunde,
Volume 7, 1912/13.
8. Die Ergebnisse der Schweizerischen Gronlandexpedition 1912/13. Zeit-
schrift flir Gletscherkunde, Volume 12, 1921122.
9. With E. v. Drygalski: Ober das Eis der Antarktis und der subantarkti-
schen Meere. Zeitschrift flir Gletscherkunde, Volume 13, 1923/24.
10. With Meinardus: Ober die hypsographischen Kurven Gronlands und der
Antarktis und die Normalform der Inlandeisoberjliiche. Zeitschrift flir
Gletscherkunde, Volume 15, 1926127.
CARTOGRAPHY
1. Bericht iiber das Projekt einer Erdkarte im Maflstab 1:1.000.000. 11.
Jahresbericht der Geographischen Gesellschaft Bern, 1891192.
2. Schweizerische Reliejkarten. 12. Jahresbericht Geographischen Gesell-
schaft Bern, 1893.
3. Die Frage der Weltkarte1:1.000.000 vor dem Londoner Geographen-
kongrefl. 14. Jahresbericht Geographischen Gesellschaft Bern, 1895.
4. Ober Karten der Volksdichte. Zeitschrift flir Schweizerische Statistik,
1903.
5. Neue Karten des Montblancgebietes. Zeitschrift flir Gletscherkunde,
Volume 2,1908.
6. Zur Frage der Farbenplastik in der Kartographie. Mitteilungen der
Geographischen Gesellschaft Wien, Volume 52, 1909.
7. Die internationale WeltkartenkonJerenz in London. Mitteilungen der
Geographischen Gesellschaft Wien, Volume 53, 1910.
8. Oberleutnant E. v. Dreis Stereoautograph als Mittel zur automatischen
Herstellung von Schichtenpliinen und Karten. Mitteilungen der Geogra-
phischen Gesellschaft Wien, Volume 54, 1911.
9. Die internationale WeltkartenkonJerenz in Paris im Dezember 1913.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 57, 1914.
10. Die Dachstein-Karte. Mitteilungen des Deutschen und Osterreichischen
Alpenvereins 1916.
nico.stehr@zu.de
324 EDUARD BRUCKNER
11. Hundert Jahre Militiirgeographisches Institut. Mitteilungen des Deu-
tschen und Osterreichischen Alpenvereins, 1919.
12. Die Entwicklung des Kartographischen, friiher Militiirgeographischen
Institutes in der Zeit des Umsturzes bis Ende 1923. Mitteilungen der
Geographischen Gesellschaft Wien, Volume 66, 1923.
BIOGRAPHICAL
1. Nekrolog auf E. Hagenbach-Bischoff. Zeitschrift fUr Gletscherkunde,
Volume 5, 1910111.
2. Dr. Josef Roman Ritter Lorenz von Liburna. Sein Leben und Wirken.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 55, 1912.
3. F.A. Forer. Zeitschrift fur Gletscherkunde, Volume 7, 1912/13.
4. Prof James Geikiet. Zeitschrift fUr Gletscherkunde, Volume 9, 1914/15.
5. Alfred Gruncl. Mitteilungen der Geographischen Gesellschaft Wien,
Volume 58, 1915.
6. Alexander Woeikof. Zeitschrift fur Gletscherkunde, Volume 10,
1916/17.
7. Prof Dr. F. Miihlbergt. Zeitschrift fUr Gletscherkunde, Volume 10,
1916/17.
8. Julius Hann t. Mitteilungen Geographischen Gesellschaft Wien, Volume
64, 1921.
9. Sven Hedin zum 60. Geburtstag, Mitteilungen des Deutschen und
Osterreichischen Alpenvereins 1925.
10. Robert Siegert. Pet.ermann' s Mittheilungen, Volume 72, 1926.
11. Prof Dlinto Marinelli t . Zeitschrift fUr Gletscherkunde, Volume 15,
1926/27.
MISCELLANEOUS
1. Uber die angebliche A'nderung der Entfernung zwischen Jura und Alpen.
11. lahresbericht der Geographischen Gesellschaft Bern, 1891192.
2. Uber die Geschwindigkeit der Gebirgsbildung und der Gebirgsabtra-
gung. Himmel und Erde, Volume 6, 1894.
3. Die Schweizerische Landschaft einst und jetzt. Rektoratsrede, given on
18. November 1899, Bern 1900.
4. Dalmatien und das Osterreichische Kiistenland. Talks given on the
occasion of the first university roundtrip. Wien und Leipzig, 1911.
5. Das Pjlanzenschaf(Baranetz}. Russische Revue, Volume 21, 1882.
6. Wanderungen des Elentiers in Ruj3land. Kosmos, Volume 1, 1884.
nico.stehr@zu.de
LIST OF PUBLICATIONS OF EDUARD BRUCKNER 325
7. Die Stel/ung der Geographie auf dem Gymnasium. Talk given in Bern
1893.
8. Ober die Heranbildung der Geographielehrer an der Un iversitat.
Geographische Zeitschrift, Volume 15, 1905.
9. Die groj3te Volkshochschule der Welt. Internationalen Monatsschrift fUr
Wissenschaft, Kunst und Technik, Volume 8, 1914.
10. Die Kriegstagung deutscher Hochschullehrer der Geographie Ostern
1916 zu Heidelberg. Mitteilungen der Geographischen Gesellschaft
Wien, Volume 59, 1916.
11. Rapport sur l'exposition de geographie scolaire. Compte rendu du V.
Congres International des Sciences Geographiques, 1891.
12. Bericht fiber den V. Internationalen Kongrej3 der geograph. Wissen-
schaflen zu Bern vom 10.-14. Aug. 1891. 11. Jahresbericht Geographi-
schen Gesellschaft Bern, 1891192; at the same time published in French
13. Der VI. Internationalen Geographischen Kongrej3 zu London 1895. 14.
Jahresbericht der Geographischen Gesellschaft Bern, 1895.
14. Bericht fiber den VII. Internationalen Geographenkongrej3.17. Jahres-
bericht der Geographischen Gesellschaft Bern, 1899.
15. Die transkontinentale Exkursion der Amerikanischen Geographischen
Gesellschafl durch die Vereinigten Staaten, August bis Oktober 1912.
Mitteilungen der Geographischen Gesellschaft Wien, Volume 56, 1913.
16. Die k. k. Geographische Gesellschafl und die Entwicklung der Geogra-
phie in den letzten zehn Jahren. Mitteilungen der Geographischen
Gesellschaft Wien, Volume 60, 1917.
17. Die wissenschafllichen Ergebnisse der Pamirexpedition des Deutschen
und Osterreichischen Alpenvereins 1913. Mitteilungen des Deutschen
und Osterreichischen Alpenvereins 1924.
18. Eine Flaschenpost vom Tegetthoff. Mitteilungen der Geographischen
Gesellschaft Wien, Volume 65, 1922.
nico.stehr@zu.de
Subject Index
Aar valley 214 Ascension 94
Aargau 175, 176 Aschur-Ade 49, 52
Abistada Lake 173 Asia 42, 62, 84, 89,110,120,272,277,
Academy in Paris 93 279,296
Actinometer 146 Astrabad49
Afghanistan 112, 173 Astrachan 49, 56,131,135,138
AJrica6O, 62, 72,73, 84,89,96 Astrological superstition 244, 245, 246
Agriculture 106, 177 Atlantic Ocean 69, 269
Air 98, 196,198,201,202,203,206,209 Atalantic states 230, 260, 262, 268
Alatau 256 Aubonne 71,150,152,155,156,158,159
Alps 4,7,8,55,58,60,64,65,74,81,92, Augusta, Ga 262
93,110,114,121,122,123,124, Australia 10, 18,42,67,48,73,77,84,
125,142,161,171,172,189,193, 95,105169,173,187,190,279
194,195,198,200,201,207,208, Austria 6,8,13,227,228,229,238,240,
211,214,217,218,222,224,280 294,210
Alsace 111 Austrian Meteorological
Alster 38, 39, 42 Observation-Network 248
America 10, 12, 15,17,42,73,89,108, Austrian Society of Engineers and
139,140,177,187,119,223,228, Architects 101, 103
230,236,237,272,273,279,285,
286,287,291,292,293,295,299 Baensch48
American Association for the Baku 49, 52, 66
Advancement of Sciences (AAAS) Balkan states 290
29 Baltic 3, 8, 47,48,49,50,51,52,53,54,
American Civil War 291 55,56,57,58,60,64,137,138,139,
American Geographical Society 4, 22 140,141,142,144,171,174,181,
299,315 182,184
American Great Lakes 220 Baltic provinces 138, 144, 174
Amur66,85,259,263,264,265,266, Baltic Sea 3, 4, 7,48,49,50,51,52,53,
277,279,301 54,55,56,57,58,60,64,171,181,
Aralsee 255 182,184
Arcadia 106 Baraba255
Archangelsk 130 Barley 237, 281
Arensburg 175 Barnaul257
Argenteuil71, 149, 152, 158 Basel 97, 179, 180,181,182,184,320
Aridity 263 Bathurst 172
Arizona 300, 303, 305 Bavaria 195,211,217
Armania 173 Bavarian Alps 195,211,212
Artlenburg 38, 39, 40, 41, 42 Bavarian Central Registry 215
327
nico.stehr@zu.de
328 NICO STEHR AND HANS VON STORCH
Bavarian Meteorological Institute 224 240,259,262,263,265,267,271,
Bavarian stations 214 280,286,307
Bear Island 80 Central Italy 66, 70
Beaune 71, 147, 151, 153, 154, 158 Central Meteorological Station in Bavaria
Belaja 130, 133, 137 217
Belgium 227, 229, 230 Changing water levels 17
Berlin 7, 21, 23, 24, 29, 30, 31, 34, 36, Chatillon-sur-Saone 149
37,38,42,44,48,50,54,63,66,81, Chemnitz 180, 181, 182, 198
82,84,90,98,99,104,113,115, Cherbourg 183, 184, 185
172,180,181,182,197,223,224, China 16,89, 113,296
234,238,239,265,273,314,320 Chur 109, 216
Bern 6, 22, 96,109,200,214,223,234, Churwalden 216
246,247,297,301,314318,320, Climate change 88, 297, 300
323,324,325 Climate Determinism 16
Biometeorology 11 Climate variability 18,287
Black Sea 3, 8, 47, 48, 50, 58, 60, 64,184 Climate variations 114, 171, 224, 236,
Bogoslowsk 56,57, 103, 133, 137 280,281,297,301
Bora winds 115 Climatic changes 63, 173
Bosporus 49 Climatic variation 23, 127, 177, 178, 190
Brandenburg 281 Climatic variations 23,127,177,178,190
Brazil 95 Cloud cover 202, 203
Bremen 29, 31, 34, 36, 37, 38, 42, 44, 59, Colorado 22, 295, 300, 303, 305
66,259,263,265,266,267,301 Columbia River 300
Brest 183, 184, 185 Columellus 147
Bruxelles259, 263,265, 266,267 Comptes Rendus 93, 99,118,183
Buenos Aires 117, 177 Conrad Maurer 109
Bureau Central Mereorologique de Cretaceous Period 63, 82, 84
France 71,109, 114, 145 Crete 106
Bureau der Baudeputation Sektion fUr Crimean Wars 233
Strom- und Hafenbau 38 Crop 176,228,233,237,281
Burgundy 147 Crops 282, 307, 308
California 260,261 Dakota 106,295,303,305
Camtschatka 85 Dalles 300
Canada vii, ix, 112 Danube 60, 65
Cape Verde Islands 177 Danzig 53
Carboniferous period 77 Delphi 244
Carpathian mountains 122 Departement Herault 148
Caspian See 3, 8, 47, 48, 49, 52, 53, 54, Departements Pyrenees Orientales 93
55,56,57,58,60,64,71,74,90, Departement du Var 93
125,161,167,168,172,296 Department of the lower Charente 93
Caucasus 56 Deviations 137
Central Europe 18,56,57,82,86,160, Dew-point 209
163,175,121,122,227,229,233, Dijon 66, 71, 147, 149, 150, 151, 153,
154, 158, 159
nico.stehr@zu.de
SUBJECT INDEX 329
Diluvial Ages 85 Forests 90, 91, 97, 98, 99
Diluvium 79,319 Forssmann 120
Dnjepr 50,131,135,138,258 Fossil 81
Don 50, 97, 258, 259, 301 Fourier components 9
Don River 301 France 19,71,93,101,111,112,114,
Donau 50, 57, 59,130,131,135,138, 115,118,127,145,146,148,156,
173,174 157,158,159,160,161,175,176,
Dove 112, 171, 187 220,225,227,229,231,235,238,
Drought 221, 286 239,240,280,282
DunaI31,134, 137,138,142, 144, 163 Frankfurt 23, 24, 26, 31, 34, 36, 38, 42,
Dusseldorf 59 44,66,82,180,182
Frankfurt am Main 24, 180, 182
East England 279 French Chamber of Deputies 93
East India 177
East of SibiriaE 66 Galicia 293
East Russia 66 Ganges 69
East Siberia 69, 259, 262, 266, 267, 280 General Assembly of the German
Eastern Alps 56, 172, 193 Meteorological Society in Karlsruhe
Ebermayer 98, 100 47
Ecoomic conditions 285, 286 Geneve 112, 122, 123, 150, 187
Egypt 73, 74, 92, 101,113,177, 178, 187 Geographical Journal 281
EIbe River 28,36,38,39,40,41,42,43, Geographische Gesellschaft in Hamburg
97, 129 3,25
Emigration 285, 293, 295, 310 German Empire 82,122,271,272,273,
Engadin210 275,276,292,293
England 66, 69,100,111,221,225,227, German Meteorological Society 25, 64,
228,229,231,232,233,235,236, 183
238,239,240,265,271,282,286, Ghats 94
290,302 Glaciers 167, 168, 171,189,315
Eocene 80, 81, 82, 85 Glommen51
Eskimos 110 Goktscha Lake 173
Europe 16,272,297,301 Grain 4,175,219,224,228,229,232,
European part of Russia 128,235,255, 235,238,239,240,290,297,301,
259,266,267 307
European Russia 69, 232, 259, 267, 301 Grand Junction 300
Evaporation 209 Grapes 147,238
Export 233, 241 Great American Desert 106
Great Britain 100, 111,221,222,271,
Famine 23 286,289,292,293,310
Finland 53,132,138,141,142,143,185 Great Salt Lake 10,60,73,106,107,108,
Floods 117, 173 171,172,177,178,295,306
Fluctuations 3, 28, 36, 44, 47, 52, 64, 65, Greece 113
72,113,183,228,229,232,255, Greenland 80, 81, 85, 86,109,110,181
256,262 Greifswald 53
Fluctuations of 64 Grinnell Island 80
nico.stehr@zu.de
330 NICO STEHR AND HANS VON STORCH
Grinnell-Land 81,85 Interior 66, 234, 242
Groundwarer3, 25,29, 30, 31, 32,33, 35, Iowa 66, 100
36,37,43,44,178 Ireland 111, 221, 222, 252, 271, 280, 286,
Groundwarer levels 43 292
Gulf states 260 Irgiz 256, 257
Gulf Stream 115 Irtysch 130, 132
Isles of Cap Verde 94
Haemus 104 Italian Alps 193
Hamburg vii, 3, 6, 25, 28, 31, 32, 33, 34, Italy 19,67,69,92,112,293
35,38,39,40,41,42,43,44,45, Japan 81, 85
112,178,180,182,317,318,321 Japanese war 293
Handbuch der Klimatologie 14,22,23,78 Jenissei 131, 132, 135, 139
Harvest 4,145,150,151,160,161,164, Jevon 171, 173, 177, 188
166,167,168,176,219,237,278, Jupiter 244, 245, 250
289,297,301
Harvesting 238 Kama 130, 131, 133, 134, 135, 137, 138,
Helsingfors 66, 175 258,259
Herault 93 Kansas 66, 106,294,295,303,305
Hochburg 175,176 Karlruhe 6, 64, 241
Hochkreuth 195,212 Katharinenburg 195,198,201,207
Hohe Tauem 171 Kempren 214, 215, 216
HohenpeiBenberg 121 Kidron 92
Holland 222, 240 Kirghiz sreppe 256
Hudson 113, 130, 131, 135, 138 Kronstadt 174
Hudson River 113 Kumo 131, 138
Hungary 93, 223, 293, 310 Kiimbach 71,145,150,152,155,158
Kiistrin 48, 57, 59, 66
Ice 77,87,117,127,130,135,137,141, Kyro 131, 137
143,166,208,209
Iceland 69,85, 110, 181,280 Lake Alakol 256
Idaho 294, 303, 305 Lake Aral 255, 256, 267
Illinois 100 Lake Bathurst 105
Immigration 309 Lake Como 193
Increase in 127 Lake Erie 175
India 60, 67, 80, 84, 94, 99,120,177, Lake Fucin 142, 167, 168, 172
188,221,236,270,286,296,301 Lake George 72, 105, 172
Indian Ocean 120 Lake Hamun 172
Indian rerritories 106 Lake Michigan 60
Indiana 100 Lake Tacarigua 95
Indus 69, 112 Lake Trasim 142
Inn 194, 195,217 Lakes 60,65, 72,167,168,172,261,299
Inn valley 194 Lakes Cowal 172
Inner Asia 60,169 Lapland 110
Innsbruck 194, 217, 218 Lausanne 71,146,150,151,152, 155,
Innthal 217 156,158,159
nico.stehr@zu.de
SUBJECT INDEX 331
Le Havre 183, 184 Mississippi 60, 66, 74,107,177,260,
Leipzig 21, 23, 63, 80, 81, 82, 83, 86, 88, 263,264,265,277,294,301,303
89,91,94,98,103,109,114,115, Mississippi delta 60, 177
119,175,185,188,268,319,324 Mississippi-valley 266, 279
Lena River 81, 200 Missouri 100, 108,260,261,299
Lindau 214, 215, 216 Mongols 296
Little Belt 182 Montana 106, 303, 305
Livingstone 96 Montmorency 149
Lockyer 120 Moon 245
London 22, 23, 24, 80, 89,94, 101, 108, Mormons 107,295
111,112,323,325 Moscow 66, 230
Louisiana 113 MountSantis 194,195,217
Lugan 56, 57, 66 Mount Splugen 193
Lund 115,240 Munchen 3, 6, 29, 31, 34, 35, 36, 37, 38,
Lustrum 54,138,139,140,141,142,161, 42,43,44,93,115,123,178,180,
162,182,272 182,195,212,213,229,238,240,
281,289,290,307,308
Mackenzie River 85
Madeira 94 Nassau 175, 176
Madras 60, 67, 185, 189,301 Nebraska 106,295, 303, 305
Magdeburg 38,39,40,41,42,57,59 Nertschinsk 60,66,259,263,266,267,
Main 23, 26, 36, 66 277,279,301
Mars 244, 245 Neufahrwasser 53, 58
Mass 14,262,285,295 Neuglietzen 57
Mass migration 295 Neusiedler Lake 65
Mauritius 67, 94, 176 Neusie1der See 8, 56, 65,124,161,168
Mediterranean Sea 48,49,89,91 Nevada 294, 295, 303, 305
MemeI53,54, 57,58, 59,97, 131,135, New Bedford 262, 263
138 New England 70, 71, 93,100,113,115,
Mercury 244, 250 260,261,262,263,264,265,268,
Meteorological Centre in Zurich 200, 251 301,302
Meteorological observations 165,280 New England states 100, 113,260,262,
Meteorological Station at Davos 200 265,268
Meteorologische Zeitschrift 14, 27, 48, New Mexico 94, 294, 303, 305
50,64,92,94,99,100,103,108, New South Wales 105, 117, 172, 173,
115,198,217,221,259,313,314, 177, 188
315,321,322 Newaa 73, 130, 131, 134, 137, 138
Michigan 30, 31, 32, 62, 66 Newhaven 113
Middle Ages 244, 270, 285 Nikolajesk 66, 259, 263, 266, 267, 279
Miesbach 195,212,213 Nile 60, 62, 73, 92, 177
Milano 118, 119, 121, 123 Norman settlements 109
Minnesota 100 North America 10,12,15,17,19,62,66,
Miocene 80, 81, 82, 85 67,73,77,80,81,100,112,138,
Miocene flora 81, 85 169,177,187,219,235,243,266,
272,277,279,286,287,289
nico.stehr@zu.de
332 NICO STEHR AND HANS VON STORCH
North American War of Independence Po delta 60
219 Poland 105, 144
North Atlantic Ocean 72, 109,221,271, Pontus 46, 50, 53, 58
280,286 Port d' Alger 92
North Dakota 294, 295, 303, 305 Praha 3,26,66,84, 120, 123, 185
North Sea 47, 58, 60 Precipitation 21,29, 123,257,258,259,
Northern Cape 221, 270 273,275,277,279,281,282
Northern France 66,158,240,270 Provence 93
Northern Gennany 66, 270, 278, 279 Prussia 19, 104, 105, 175, 176, 122, 123,
Northern Lights 221 225,227,228,229,237,240,255,
Northern Russia 49, 138, 174,221,270 271,281,287,288,289,307,308
Norway 181,286 Prussian 56
Nyassa 72, 173 Public University Lectures 7
Pully 71,146, 150, 155, 156, 158
Oats 237, 281
Obersulzbach glacier 122 Rain 29, 35, 43, 62, 70, 92, 118,238,239,
Ocean 24, 301, 303 240
Oder 51,57,59,65,66,96,102,104, Rainfall 30, 36, 57, 60, 68,69,77,93,94,
173,183,276 99,107,108,161,162,227,228,
Office for Marine Weather (Seewarte) 6 229,231,241,256,257,260,261,
Ohio 66,100,175,176,227,230,231, 262,275,287,290,291,293,302,
242,261,263,264,265,267,277, 307,308,309,310
279,288,289,301,308 Rainfall fluctuations 231
Ohio Valley 261, 264, 265, 217, 301 Reflection 197
Oka 97, 258, 259 Rhein 59, 96, 97,102,103,173,176,273,
Onega 131, 133, 134, 137 275,276
Orsova 57, 59 Rhine 57, 65,102,103,193,217
Oscillation 14,22,55,120,127,159,165, Riga 66, 132, 134,136,137,138,142
263,267,300,301,302,303,305, River 56, 72, 73,97,102,105,135,138,
306,310,315 141,143,166,173,174
River levels 38, 39
Pacific coast 260,301 River navigation 73
Padua 188 Rocky Mountains 81, 260
Palastine 113 Rosenheim 212,213,214,215,216
Parana river 177 Royal Bavarian Meteorological Centre
Paris 57, 59, 66, 71,89,96,99,101,109, 195
111,113,119,145,147,160,183, Rugen 182
239,323 Russia 6,17,18,19,56,60,68,97,98,
Pernau 175 127,128,137,138,139,140,141,
Perpignan 149 142,143,174,195,219,221,222,
Persia 89, 172 223,224,225,226,227,228,230,
Phenomena 62, 119, 175 231,232,233,234,235,236,241,
Phoenix 300 242,256,258,259,266,271,280,
Phylloxera 146 282,286,288,289,290,293,297,
Pierrefeu 149 301,308,310
nico.stehr@zu.de
SUBJECT INDEX 333
Rye 228, 229, 237, 239, 240, 241, 281 St. Helena 94
St. Petersburg 56, 57, 62, 66, 73, 110,
Sachalin 85 117,118,129,130,131,137,138,
Sagastyr 200,201 174,235
Salt Lake City 300 Stations 66, 67, 130, 132, 135, 151, 152,
Salzach 6, 217 153,154,155,156,158,237,261,
Salzburg 31,34,35,36,42,43,44,178 262
Santa Fe 177 Steppe 256, 257, 267
Saturation 29, 30, 35, 36 Stolpmunde 53, 58
Saturn 244, 245, 250 Stralsund 53
Saxony 281 Strzelecki 96
Scandinavia 112,221,222,265,270 Stuttgart 22, 23, 24, 71, 78, 89,92,93,96,
Schott el Djerid 90 102,119,123,145,150,152,155,
Schweinfurth 74, 101, 187 158, 189
Scotland 66,69,110,111 Sulphur 148
Seine 57, 59, 65,152,173,184,188 Sun 62, 245
Settlement 4, 299 Sunspots 119
Shetland Islands 11 0 Sweden 53, 185,286
Shipping 175 Swinemunde 50, 51, 54, 57, 183
Siberia 10,68,69, 73, 74, 94, 110, 127, Swiss Alps 81, 193,216
128,137,138,139,140,141,142, Swiss Department of the Exterior 235
143,144,174,176,178,186,187, Swiss Meteorological Centre 199,250,
225,263,264,265,280,301 252
Sigillaria 84 Swiss Meteorological Centre in Zurich
Signal-Service 107 199
Sound 58,100 Swiss Society of Natural Science 249
South Africa 84, 96 Switzerland 15,71,80,86,93, 101, 110,
South American Andes 194 114,127,145,149,156,157,158,
South East Russia 174 159,161,175,176,193,211,214,
South West Germany 145, 158 222,225,232,239,240,246,247,
South West Russia 144, 174 271,289,307
Southeast Russia 138, 219, 279 Syria 113
Southern Europe 222, 271, 286
Southern France 158,222 Tanganyika 72, 173
Southern Germany 28, 29,160,175,176, Tarbagatay 256
223,232,240,289 Tarim basin 296
Southern Pacific Company 306 Tauern 171,217,315
Southern Russia 221, 222, 233, 271, 279, Temperature 48, 70, 77, 112, 113, 114,
286 118,119,123,129,160,161,162,
Spitzbergen 80, 81, 85, 86, 181,320 199,201,202,203,206,208,209,
Spree 36 213,215,221,224,269,270,272,
Ssuchona 131,137 287
Ssyssola 130, 131, 133, 134, 135, 137, Tertiary Period 80, 82, 85, 86, 87
138 Texas 106, 303, 305
St. Cruz 95 Textbook on Agriculture 146
nico.stehr@zu.de
334 NICO STEHR AND HANS VON STORCH
The Limping Messenger 246 Virginia 113
The Swiss Farmer 246 Volga 49, 56, 65, 223, 296
Thunderstorms 247 Volger27
THlis 56, 57, 66 Volnay 151, 154, 158, 175, 176
Tilsit 57, 59, 66,130,135
Toulon 149 Wamemiinde 53, 54
Traunstein 214, 215, 216 Washington 73,83,100,106,107,111,
Travemiinde 50, 53, 54 177,260,261,300
Trevandrum 188 Water volume 62, 104, 128
Tropics 271 Weather conditions 178
Trottier 92 Weather predicitons 243, 250
Tsad 173 Weichsel57, 58, 59,65,97,105,131,
Tschussowaja 131,135,138 135,138,173,174,276
Tunisia 90 Wendelstein 195,212,213
Tuscany 112 Weser 36, 57, 59, 65, 97,103,173
Typhus 3, 25,29, 30, 31, 32, 33,34, 35, West Indies 95, 177
39,40,43,178,179,180,182,321 West Russia 66, 144
Tyrol Alps 198 West Sibiria 66
Wheat 176,228,229,237,238,239,240,
u.S. Weather Bureau 260 241,281,290
Union Pacific Railroad 306 Wheat prices 290
United States 4,10,15,17,18,22,24,64, White Sea 137, 221, 270
66,69,81,100,106,107,112,113, Wiek53
119,230,241,252,260,261,262, Wien 3, 4,13,21,22,23,24,47,56,65,
266,267,280,282,285,286,287, 66,72,77,82,83,86,89,90,91,
289,290,291,292,293,294,295, 98,101,109, 115, 118, 119, 123,
299,301,303,304,305,306,308, 124,185,194,196,223,224,239,
309,310 255,272,285,297,300,301,315,
United States of America 15,106,280, 319,320,321,322,323,324,325
285,287 Wilson Flagg 91
Urru56, 138, 139,140, 143, 174 Wind 206
Uranus 250 Wismar 53,184
Wjatka 131, 134, 138
Variations 3, 4,49, 52,64, 65,68, 70, Wolga 97,102,131,134,135,138,258,
107,113,114,122,127,131,138, 259
141,143,145,150, 161, 162, 163, Wologda 131,133,134,137
165,167,168,171,176,178,237, Wiirm46,47
238,239,240,241,255,258,267, Wiirttemberg 155, 158, 175, 176
272,275,277,293,301 Wyoming 295, 303, 305
Variations in 4, 64,150,171
Vendee 93,115 Ziiricher See 243
Venezuela 95
Venus 244, 250
Verdun 71,148,149,152,158
Veytaux 71, 150, 158, 159
nico.stehr@zu.de
Name Index
Abbott 105 Capre 243, 249, 250
Abels 198 Celoria 119, 120
Adams 107 Chambers 111, 120
Aeolos243 Charpentier 86, 87
Agassiz 97 Chavanne 89,117
Anderlind 92, 101 Cicero 244
Angot71, 114, 127, 145, 146,147,148, Qave91
149,159,160,175 Coaz 110
Apollo 244 Columellus 147
Arago 91,109,111,112,113,114 Croll 87
Archibald 96, 120 Curtis 107, 108
Aristotle 2, 244 Czemy 109, 110
Arrhenius 12
Darwin 177
Baensch48 Davy 101, 171, 188
Baxendell 119 Dawson 120
Becquerel 96, 199 de Bellevue 93
Bellew 112 de Brahe 101,245
Berg 16,20,21,256 de la Grye 171,183,184
Berghaus 92, 93, 96, 97,102,103,106, deTours 147
110 De Gasparin 146
Bemsmann96 Denison 107
Bidin 94 Denza91.92
Biot 113 Dorsey 107
Blanford 77,89,94,99,119,120,190 Dove 112, 171, 187
Blanqui 94 Draper 100, 113,130
Blavier 115 Drought221,286
Blodget 107 Dufour 77, 110, 114, 116, 145
Blytt 87 Duponchel118
Bohm96
Bourlot 111 Ebermayer98,100
Boussingault 95, 96 Ekholm 207
Brocldesby 120 Ellis 224
Broun 120 Emmerich 57, 59
Brown 188 Engels 16
Brumham 117 Engler 77, 80 81, 82, 85
Buchan 111 Falb 243, 246, 247, 248, 249, 250, 252
Burton 101 Fautrat 91,98,99
Fechner 168
335
nico.stehr@zu.de
336 NICO STEHR AND HANS VON STORCH
Filipof 49, 52, 55 Helmersen 97, 102, 103
Fischer 24, 63, 77, 88, 89, 90, 91, 112, Hendorf of Wittenberg 245
189,322 Henry 260, 261, 262, 263
Forel64, 74, 78,122,123,124,125,127, Herder 16
146,189,224,324 Herschel 91
Forry 113 Hill 119, 120
Forssmann 120 Home 91, 92, 94, 95, 96
Friedrich 16, 207 Honsell 62, 104
Friesenhof 246 Hood 306
Fritsch 96, 119, 224 Hornstein 118, 120
Fritz 56, 62, 74, 95,119,120,124,171, Humboldt 89, 95, 113
175,176,177,188 Hunter 120
Huntington 7, 9,16,23,296
Gardner 81 Hypocrates 244
Gautier 112, 119
Geering 232, 235, 236 Ideler77, 106, 110, 111, 112, 113, 115
Gessner 101 Ignatow256
Gibson 94 Imhof 200
Gilbert 73, 74, 106, 107, 177, 178 Iris 243
Gladbach 243, 249, 250
Glaisher 111 Jager 246
Gould 117 Jamieson 100
Grager92 Japan 81,85
Grebenau 101, 102, 104 Jaovns 120
Greely 107 Jeffe~on 113, 115
Gru6115 Jelinek 120
Gwilliam 117
Kalm 113
Hagen 56,103,104,171,188 Kamtz 91,100
Hahn 119, 120 Karsten 182
Hamilton 111 Kasthofer 93,110
Hann 12, 13, 14, 15, 16, 18,20,22,23, Klug 188
24,74,78,96,99,103,119,171, Knauer 246
186,187,203,260,280,287,319, Knoch 14, 15,16,22,23
324 Koppen 6, 14, 16,70,77,97, 102, 118,
Harrington 107, 108 119,163,176,200,315
Hartt 95 Krafft 77, 117
Hazen 107, 120 Kremser280
Heer 77,80,8182,83,85,86 Kurzbracke 57, 58, 59
Heim 189
Heinrich 96 LaCour 101
Heintz 224 Lacey 11,13
Heintze 259 Ladoucete 93, 111
Hellmann 99, 245 Lamb 12, 16,20,23
Hellpach 13, 23
nico.stehr@zu.de
NAME INDEX 337
Lang 7,8,36,58,60,64,65,74,78,123, Neumayr77,81,83,84,85,86
124,189,224,243,246 Newton 113
Larochefoucauld-Liancourt 112 Niel92
Lecoq 91 Noah 113,245
Lendenfeld 105 Nordenskjold 80,81
Lenz55
Leresche 110 Odart 146
Lespiault 115 Overzier 246
Lesquereux 81
Lickscha 131, 138 Partsch 77, 90, 113
Lindau 214, 215, 216 Paschen 171, 184
Livingstone 96 Penck 5, 6, 28, 43, 66, 81, 82, 84, 87, 178,
Lockyer 120 196,315,317,318
Longmann 11, 23 Pernter248
Loomis 113 Pettenkofer 26,27,28,30,33, 178
Lorenz 184,324 Pfeil 97, 102
Lornezonil71,188 Piazzi 119
Luther 245 Picot 111, 112
Lyell 12,77,79 Pilgram 163
Plantamour 112,171,187
Marchand 96 Ploetz 13,23
Mares 148 Poey 120
Maria Theresia 93 Powell 73 106, 108, 177
Markham 103 Pralle 104
Marmont92 Pully 71,146,150,155,156,158
Marsh 91, 93, 94
Marti 243, 249, 250, 252 Ratze1198,285
Martins 111, 112 Reincke 25, 28,38
Mathieu 91, 99 Reis 188
Maurer 109 Reiss 194
Maydell 48, 50, 58 Renou 77, 118
Meldrum 94,119,120 Richter 8, 64, 74, 78,122,123,171,172,
Merian96 189,193,224,315
Meyer 48 Riviere 93
Michelier 109 Roscoe 119
Mohn 181 Rosenheim 212, 213, 214, 215, 216
Montesquieu 16 Riihlmann 120
Morrow 107 Russel 117, 173
Mosca 200 Rykatchev 130, 142
Miiller 31, 40, 44, 207
Muret 110, 316 Sachs 95
Murphy 92, 94 Saporta81
Sartiaux 98
Nares 80 Sasse 102
Nathorst 77,80,85,86 Schlichting 102, 103, 104, 105
nico.stehr@zu.de
338 NICO STEHR AND HANS VON STORCH
Schmick 87,89, 114 v. Czemy 86, 119
Schmid 104, 105, 190 v.Kemer217
Schmidt 98, 110 v. Middendorf 110
Schott 90, 100, 113, 118, 119 v. Sonklar78, 121, 122, 123
Schreiber 198, 268 v. Wagner 103
Schulz 277, 278, 279, 281 Van Bebber 93
Schwabe 119 Van den Brinken 94
Seibt 48, 50, 54 Vrrgil224
Seneca 104 Volney 113
Sieger 65, 72, 74, 173, 178, 224, 265, vomRath 91
321,324 von Humboldt 95
Sievers 173
Simony 91, 93, 95,115 Webster 113
Slutsky 9 Wehrli 127, 146
Smith 119 Weilenmann 119
Snow 107, 194 Wex 51, 77, 97, 98,100,101,102,103
Sokolof55 Wheeler 94
Sombart 13, 24 Whipple 120
Soyka 26, 27, 28, 33, 34, 36, 37, 42, 44, Whitney 15,24,63,77,80,88,89,91,93,
178 95,96,103,108,109,110,111,112,
Stein 296 113, 114, 115, 189
Stewart 119, 120 Wild 60, 103, 118, 129
Stamer 245 WIlliamson 113
Stone 119 WIlson 91, 96
Strabo244 Woeikof49,50,60,62,99, 130, 132, 196,
Strachey 120 199,205,206,207,216,221,255,
Strzelecke 96 256,257,267,324
Studer 145 Wolf 108, 119, 120
Stiibel194
SueS 171,185 Zenker 197
Swarowsky 8, 56, 64, 65,124,125 Zeus 263
Symons 117, 120 2JmmermannI12,119,171,187
Tacitus 92, 244
Tangniliew 232, 256
Theophrast 104, 106
Thomas 107
Thomassy 113
Thompson 107
Tsar Nicholas 97
Tschudi 110
Umlauff16
v. Baeyer48
nico.stehr@zu.de
