The Western aqueduct of Nicopolis ad Istrum is one of the best studied Roman conduits excavated so far in modern-day Bulgaria (Църов 2017, 165–325). The focus of this article is the water-catchment at the village of Mussina, excavated by...
moreThe Western aqueduct of Nicopolis ad Istrum is one of the best studied Roman conduits excavated so far in modern-day Bulgaria (Църов 2017, 165–325). The focus of this article is the water-catchment at the village of Mussina, excavated by the author in 2017. Furthermore, personal observations along the route of the aqueduct are also presented here. The water-catchment is situated in the Polyanata site, 587 m southwest of the church in the village of Mussina. It is located about 30 m northeast of the entrance to the Mussina Karst Cave and 1.50 m north of the remains of the Raevi Mill (Fig. 1, Ill. 1). After it was built the terrain was re-shaped (Fig. 12). North of the catchment there is a modern dam,
built in the late 70s of the 20th century along with a trout hatchery (Fig. 1, Fig. 12). This raised the terrain in front of the spring-box, and nowadays it is partially or completely underwater (Fig. 9, Fig. 15, Ill. 1, Ill. 2). It is an octagon-planned structure made up of 32 stone blocks of dolomite limestone, as is the surrounding rock massif extending to the
Roman stone quarries between the villages of Hotnitsa and Samovodene (Appendices 1and 2). The inside width of the facility on the second row is 4.23 m between the long sides of the blocks and 4.60 – 4.65 m between the corners. Its outer width is about 5.30 m. The total height of the stone masonry is 2.20 m and the wall thickness is 0.50 – 0.55 m. The blocks are arranged in four vertical rows of 8 pieces per row (Fig. 6, Fig. 7, Fig. 8). The stone blocks are numbered from 1 to 32 starting from the bottom row (row 1).
The water comes from the cave (Fig. 3, Fig. 4, Fig. 5). There are three artificial dams inside it. Two of them were described by D. Tsonchev as the “Big Dam” (Fig. 4) and the “Small Dam”, and the author noted that the smaller one had a mortar with crushed pieces of bricks in its structure and was probably from the Roman period (Цончев 1935, 134). The 2017 site observations indicated that the big one was built entirely of stones
and modern bricks. The structural features of the Small Dam do not provide grounds to refer it to a certain period. However, after sacks of straw were piled over it to divert the water to the entrance of the cave, it was found that the water level at the catchment did not change. This means that the Dam was built to divert water to Branch III (after Цончев
1935, 134, 137), which supplied water to the mill in front of the cave. Therefore the Small Dam seems to have been built in modern times. In a description of the cave made during the mapping expedition in 1978, the authors mentioned that the millers had made the Big Dam and the Small Dam around 1920, having found a previous wall of an older mill (Даалиев, Стрезов 1978). This hypothesis finds support in the existing third dam, made of modern bricks and concrete. It blocks Branch III (after D. Tsonchev) (Fig. 3, Fig. 5). Branch IV is presented in the cave plan, made by D. Tsonchev, as leading right to the water-catchment beyond the cave (Fig. 3). In fact, in this branch there are three in108
surmountable cracks in the rock. On the left there is a crack with underground streams that are almost inaccessible. The water passing through it probably comes from the small pond beyond the second (small) dam, and seeps through cracks in the rock. When the
water passes through the two dams it drains into the floor of hall “g” (after D. Tsonchev) and connects again with the already mentioned underground stream and takes out water of constant temperature of 13.9 – 14.1°. During the 2017 excavations it was found that the flow rate at the water-catchment was about 50 l/s. After going out of the cave, the water passes through a rock-cut underwater gallery about 1 m in width, about 3 – 4 m in length and about 1.50 m high. It is located just southwest and west of the water-catchment and is invisible on the ground surface now (Fig. 6, Fig. 10). Blocks nos. 1 and 2 on the lower (first) row of the catchment have collapsed and a joint was formed through which one can pass inside the gallery (Fig. 9, Ill. 2). A wall was also found in the gallery (Fig. 11), adjoined along the outer faces of blocks no. 1 on the first row and nos. 9 and 10 on the second. As block no. 1 has already subsided, the wall now seems to be “hanging out”. It is built of bricks, mortar, stones and gravel. Its front surface is smooth and its lower surface – irregular. The rock gallery and the wall might be a bypass to divert water during the construction of the catchment and, if needed, to drain the inside of the facility. The subsidence of the first row in the western part of the catchment shows that the bottom in this part is at a lower level. This can be explained by the fact that water from the gallery enters the catchment at this point. After the facility was abandoned and filled with trash and stones, the water burst through the structure itself and caused the blocks to be displaced.
The water-catchment is built of 32 stone blocks made of dolomite limestone, as is the surrounding rock massif (Fig. 6, Fig. 7, Fig. 8, Fig. 9, Ill. 1, Ill. 2). The blocks are arranged in four vertical rows of 8 pieces per row. A numbering of 1 to 32 is presented starting from the lowest row (row 1) (Appendix 1 – 2). All the blocks (no. 1 to no. 16) from the
first and the second rows are preserved;blocks nos. 19 – 23 and the left half of block no. 24, which is not in situ now, are preserved from the third row, and blocks nos. 27 – 30 of the fourth row. The first and the third rows are made of blocks of isosceles-like trapezoid shape, and the second and the fourth – of hexaagonal ones. The corners of the trapezoidal blocks lie on the angles of the hexagonal ones. The fourth row is drawn 0.12 m inwards the catchment and its corners are slightly eccentric compared to those of the lower rows. Perhaps a corbel arch dome was made in this way. Beds for U-shaped iron clamps, now
missing, were found on the first and second rows. Two mouths are formed in the wall of the facility. One of them is at a lower level and the main channel of the aqueduct begins from there. The other one is higher and probably served as a spillway channel, through which, in higher waters, the excess water was directed to the main channel (Fig. 2). The
following structures of eth aqueduct were localized to the northwest of the catchment: a linear rock cutting, a round hole, and part of the right wall of the channel with a buttress (Fig. 124, Fig. 14, Fig. 15).
By analogy to the water-catchment of the main aqueduct of Ratiaria, the caput aquae of the Western aqueduct of Nicopolis ad Istrum could be dated to the same period – the early 2nd c. AD, as Roman bricks with stamps of Hadrian (117 – 138) were found in the conduit structures of the aqueduct in Ratiaria. The two water-catchments are almost similar in terms of construction techniques and plans. The construction of 109
the Western aqueduct of Nicopolis ad Istrum began most probably during the reign of Trajan (98 – 117) and was continued by his successor Hadrian (117 – 138) (Църов 2017, 165 – 166).
The author of this article also documented with GPS points of the route of the Western aqueduct of Nicopolis ad Istrum (Appendix 3). The study covered the distance from the water-catchment to the water reservoir before the city. The exploration was based on points previously known from the publications, and on new ones, suggested by local guides in the villages of Mussina, Dichin and Vodoley. A total of 59 points were documented (Fig. 16, Fig. 17, Fig. 18). The estimated length of the conduit is 20.3 km. The elevation difference is 24.404 m, measured at the bottom of the channel in Mussina and the bottom of castellum aquae before Nicopolis ad Istrum. If the aqueduct was mouted 4 meters from the bottom of the reservoir, the difference in the level would be 32.404 m (Fig. 17).