Yurta‑Stroyno Archaeological Project. Preliminary Report to the Final Season of 2016

STUDIA HERCYNIA XXI/2, 99–112 Yurta ‑Stroyno Archaeological Project. Preliminary Report to the Final Season of 2016 Petra Tušlová– Barbora Weissová– Stefan Bakardzhiev ABSTRACT The last field season of the Yurta ‑Stroyno Archaeological Project conducted in 2016 focused on finishing the excavation of the five ‑room house previously discovered in the southern part of the Roman settlement. Si‑ multaneously with the excavations, atargeted field survey took place in selected areas of the rural settlement. Moreover, the material uncovered over the three years of the project is being gradually processed, some of the preliminary data of arather statistical character are also presented here. KEYWORDS Bulgaria; Roman Thrace; rural settlement; excavation; field survey; pottery. INTRODUCTION The last field season of the three ‑year joint project of the Regional Historical Museum of Yam‑ bol and the Institute of Classical Archaeology in Prague, excavating the Roman rural settlement of Yurta ‑Stroyno in south ‑eastern Bulgaria, took place in autumn 2016. From October 7th to November 13th students of Classical Archaeology1 and personnel of the Museum of Yambol2 accompanied by asmall group of local workers participated in the project. The field work was divided between two groups. One of them finished the excavation of ahouse discovered two years before, the other one focused on afield survey of selected polygons placed over the remaining settlement. Additionally, from August 27th to September 16th 2017, ashort material documentation ses‑ sion took place to finish the post ‑processing of the uncovered material, which almost doubled last year when the excavation and the field survey were taking place simultaneously. 1 Besides the two co ‑authors of this paper, students from the Institute of Classical Archaeology at Charles University in Prague took part in this season: Věra Doležálková, Viktoria Čisťakova, Jakub Havlík, Hana Havlíková, Josef Mareš, Adéla Minaříková and Anna Peterková, as well as one student of the Institut für Archäologische Wissenschaften, Ruhr ‑Universität Bochum, Clarissa Hauben‑ thal. 2 Besides the co ‑author of this paper, Todor Vulchev and Miroslav Kozarev. 100 STUDIA HERCYNIA XXI/2 Fig. 1: The overview map of the Yurta ‑Stroyno settlement with marked areas of the excavation and the field survey (by B. Weissová). THE EXCAVATION THE HOUSE Trenches 110E–100N and 110E–105N The excavation work in this area focused on disclosing the ground plan of the house located in the south ‑western part of the settlement (Fig. 1). Within the previous two years, the stone foundations of afive ‑room structure were excavated, although the course of the walls in its western and eastern ends remained unknown (Fig. 2; Tušlová – Weissová– Bakardzhiev 2014; Tušlováetal. 2015). The stone foundations of the westernmost room of the house stopped showing about one meter away from the nearby Dereorman River, running along the south/south ‑western part of the settlement. The meandering/flooding river seems to be the reason for the absence of the corner, taking down the stone foundations, made without mortar binding. Consequently, it was not possible to follow the course of the house in this direction. However, we may suppose this was originally the last room of the house, simply because of the proximity to the river which does not leave much space for its further continuation. 101PETRA TUŠLOVÁ– BARBORA WEISSOVÁ– STEFAN BAKARDZHIEV Consequently, the fieldwork focused on uncovering the house foundations in the eastern part of the structure, soon revealing the other missing end– right after the fifth room. What was, however, surprising, was the continuation of the eastern wall [SU076] further north, suggesting the northern part of the house contained acourtyard, possibly surrounded by awall (Fig. 2; Pl. 4/1). The wall running north [SU083] is preserved until alength of 7.50 m, at which point its course is disrupted by atreasure hunters’ ditch 1.60 m long and 0.90 m wide. The ditch was dug through the stone foundations which were completely displaced and thrown to the sides. Within the timeframe of the excavation we cleaned another 0.80 m of the area north of the ditch in the possible course of the wall [SU083]; no solid structure was found (Fig. 3). Based on these observations, we may suppose that the wall running north [SU083] ended within this disrupted area. While cleaning the surroundings of the wall running north [SU083], asmall size wall [SU085] (max. width 43 cm, preserved length 55 cm) appeared, running east to west– in par‑ allel with the northern foundation wall of the house [SU074]. Acollapsed layer of secondarily fired mud bricks was found by the northern side of this small size wall [SU085], it is well preserved in situ and clearly visible in the western profile (Fig. 3). Individual mudbricks bear the imprints of wood (Fig. 4), confirming the previously suggested construction technique of the house: rubble stone foundation with masonry of wood ‑clay construction. The relation of the wall running north [SU083] and the small size wall [SU085] is unclear. What we can say from the observation is that the north running wall [SU083] has amore solid structure, composed at its northern end of two rows of stones, ca. 30 cm deep and ca. 55 cm wide.3 The small size wall consists of one row of stones ca. 15 cm high and 43 cm wide. The distance between them (ca. 85 cm) seems to be intentional as there are no marks of these two structures being connected. We might be dealing here with asmall open entrance to the courtyard, from which it was possible to enter the house itself. AREA NORTH OF THE HOUSE 100E–110N SE This area was chosen for excavation as it seemed not to be affected by the treasure hunters. It is located directly north of the square excavated the previous year (100E–105N NE), which brought to light aca. 50 cm thick layer of topsoil followed by acompact grey 35–40 cm thick cultural layer filled with heterogenous archaeological material and stones, resting on the yel‑ low sandy layer [SU007] (perhaps virgin soil; Pl. 4/2). The cultural layer seems to be originally covering the whole surroundings of the house, but archaeologically can be detected only on the places not affected by the treasure hunters’ activities– which are few. The layer was laid down within asingle action as numerous pottery fragments from different heights of this layer and areas of the excavation joined together. Thus, it was preliminarily interpreted as a‘levelling’ layer composed of the settlement waste (see Tušlováetal. 2015). 3 This is, however, much less than on the southern end of the same wall [SU083], where it intersects the northern foundation wall of the house [SU074]. In this corner, the foundations of the wall [SU083] reach adepth of more than 65 cm. This difference seems to relate to the position of the house on agentle hill sloping from north to south (where, of anecessity, there are consequently deeper foundations than on the north). 102 STUDIA HERCYNIA XXI/2 103PETRA TUŠLOVÁ– BARBORA WEISSOVÁ– STEFAN BAKARDZHIEV Fig. 2: The excavated area (2014–2016) with ground plan of the house (digitalized by M. Kobierská). 104 STUDIA HERCYNIA XXI/2 Fig. 3: Photo of the northernmost area of the excavation with the interrupted north running wall [SU083], robbers’ trench, and the small size wall [SU085]. Photo by P.Tušlová. Fig. 4: Fired mud brick with imprints of wood; found north of the small size wall [SU085]. Photo by P.Tušlová. 105PETRA TUŠLOVÁ– BARBORA WEISSOVÁ– STEFAN BAKARDZHIEV The new trench has the dimensions of 2.50×2.50 m, with amaximal depth of 0.80 m and astratigraphy repeating in the already known pattern mentioned above. From the abundant material, the biggest amount of the finds belongs, not surprisingly, to the pottery. Animal bones are also quite common, even more than the architectural ceramics. Glass is represented mostly by vessel body fragments, otherwise two beads and one piece of window glass were found. From the metals, except one bronze strip, everything is iron, with the highest representation of nails (13 pcs.). Further aneedle, different components of architecture such as ahook, clamp, gusset and otherwise unidentified pieces were found. All the terracotta fragments belong to lamps and the one worked bone is a9.5 cm long pin (Tab.1). The material is in general very fragmented. Occasionally, there were found bigger parts of animal bones, or several pottery fragments from one vessel, which obviously broke on the spot when they were deposited. Tab.1: Overview of the material represented in the ‘levelling layer’, trench 100E–110N SE. The first horizontal row shows the quantity in pcs. (q.) and weight in grams (w.); the second row marks the ratio of the individual materials (based both on count and weight) within the overall layer amount. Tab.2: Overview of the represented pottery classes in the ‘levelling layer’, trench 100E–110N SE. The first horizontal row shows the quantity in pcs. (q.) and weight in grams (w.); the second row marks the ratio of the individual classes (based both on count and weight) within the overall layer amount. The pottery covers arepresentative selection of different classes (Tab.2), with the highest amount of table ware (including both fine and the so ‑called common ware, since the differenc‑ es between these two groups are not obvious in the Yurta ‑Stroyno assemblages). Asurprise in the pottery diversity is the missing storage ware, in general little represented in the excavated area, however, well known from in the field survey assemblage (see Tab.4). 105E–105N north of the wall [SU074] To rule out the possibility that the wall running north [SU083] is in fact the eastern side of another room, it was necessary to explore the northern continuation of the eastern ‑most Room E. For this purpose, part of the square north of the wall [SU074] was opened (ca. 3×2 m). No wall continuation was revealed, and the only stratigraphic unit encountered was the ‘level‑ ling’ layer. One peculiarity was abig rectangular stone (ca. 60×45×20 cm) placed within the Material overview for trench 100E–110N SE, the ‘levelling layer’ Pottery Frgs. Animal Bones AC Frgs. Glass Frgs. + SF Metal Frgs. + SF Terracotta SF Worked Bone SF Total Amount q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) 1769 17924 500 3224 410 2300 130 137 36 504 8 – 1 – 2854 24089 62% 74% 18% 13% 14% 10% 5% 0.5% 1% 2% 0.5% 00.5% 0 100% 100% q. = quantity (pcs.); w. = weight (grams); AC = architecture ceramics (tegulae, imbrices, bricks); SF = small finds Pottery overview for trench 100E–110N SE, the ‘levelling’ layer Table Ware Coarse Ware Amphorae Hand Made Grey Ware Storage Total amount q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) q. w. (g) 1135 10987 338 2673 135 2371 135 1532 26 361 0 0 1769 17924 64% 61% 19% 15% 8% 13% 8% 9% 1% 2% 0% 0% 100% 100% q. = quantity (pcs.); w. = weight (grams) 106 STUDIA HERCYNIA XXI/2 layer right in front of the wall SU062 which is dividing the Rooms D and E (Fig. 5). It has no traces of usage, but its position and height do not seem to be accidental, since it must have been protruding out of the layer. It could have, in fact, accommodated awooden (?) column holding aporch over the entrance to the house. Fig. 5: Photo of the area north of the house with the rectangular stone in front of RoomD. In the thick ‘levelling’ layer [SU082] two Severan period coins were found. Photo by P. Tušlová. Since the stone foundations bear no marks of thresholds, we cannot confidently say where the entrance to the house was positioned. However, considering the different appearance of Room D– with smaller proportions than the other rooms, pebble paving with ahalf ‑sunken terracotta tube (for the description of the situation see Tušlováetal. 2015, 248) and the newly revealed stone placed in front of the house, we incline to interpret this room as an entrance to the house. It would divide the house into two areas– three rooms at the rear of the house, near the river, facing the courtyard (living quarters?), and one room at the corner, facing the courtyard as well as the outer area– perhaps astreet (ashop or craft workshop?). Similar architecture is known from military vici and urban areas as well, where the so ‑called ‘strip houses’ (‘Streifenhäuser’) feature living quarters and rooms/aroom for crafts, agarden on one side of the house with the other side facing the street (Sommer 1998, 45–46). Within the ‘levelling’ layer, two coins were found this year placed at aca. 20 cm depth differ‑ ence. The first one (SF16_143) is of Caracalla issued in Traianopolis (198–217 AD, 214.92 m.a.s.l. Pl. 4/4:1), the second one (SF16_185) of Septimius Severus with the issue place unreadable (193–211 AD, 214.72 m.a.s.l. Pl. 4/4:2).4 Chronologically, they match well with the coin of Diadumenian (217–218 AD) found the previous season within the house foundation trench and acoin of Julia Domna (193–211 AD) 4 Adetailed description of all the coins from the project is being prepared for the final report by Petra Janouchová. 107PETRA TUŠLOVÁ– BARBORA WEISSOVÁ– STEFAN BAKARDZHIEV found in trench [SU057] next to the foundation wall (Tušlováetal. 2015, pl. 10/2:1–2). In to‑ tal, we have found six Roman coins within the three years of excavation, and the prevailing amount (four out of six) are of the Severan period. Their presence in the contexts relating to the house and the ‘levelling’ layer suggests that these two structures were most likely contem‑ porary with each other, based on the coins, erected on/after the turn of the 2nd and 3rd c. AD. E12 NW– FAS In the north part of the site, in the area which did not seem to be disturbed by the treasure hunters, asmall trench (1.2×1.2 m) was placed in the NW corner of the square E12, just one meter south of the peg of square D12. Its original purpose was to gain data for phosphate analyses, however, it also offered acomparative and stratigraphic sample for the excavation area from which it was about 130 m away (Fig. 1). The trench was excavated in eight mechanical layers, each of approximately ten centimetres thickness. The soil of the first ca. 50 cm was of adark ‑brown colour (topsoil), then it changed to arather grey, harder and compact layer. Aquantity of small ‑size gravel appeared at ca. 60 cm and gradually grew until it reached the compact stony layer at ca. 80–90 cm (Fig. 6). The trench in general yielded asmall amount of material which was fragmented and worn, mainly composed of pottery shards with few architectural ceramics, glass and metals (Tab.3). The highest amount of material– 18 pcs.– was spotted within the last mechanical layer FAS08 which was also dense in gravel and small size stones. Asmall iron nail, often considered ahobnail of aRoman shoe, was found among the gravel within the last mechanical layer. We may suppose this could be awalking surface which cannot be characterised any closer based on this small sample. However, its character– topsoil followed by ahard, compact grey layer accumulating stone gravel– is similar to the northern part of the excavated area (north of the foundation wall), although the amount of material is dramatically smaller. Fig. 6: Small trench (1.2×1.2 m) in the square D12 on the northern perimeter of the settlement at its maximal depth (90 cm). Photo by P.Tušlová. 191PLATES Pl. 4/2: Northern section of the trench 100E–110N SE with marked layers. Pl. 4/3: Selection of finds (both excavations and survey). Bronze coins: 1: of Caracalla issued in Traia ‑ nopolis (198–217 AD), from excavations; 2: of Septimus Severus (193–211 AD) the issue place is unre‑ adable, from excavations; 3: of Tranquillina (241–244 AD), minted in Anchialos, from field survey. 4: Dog ‑shaped head of abone pin, field survey. 192 STUDIA HERCYNIA XXI/2 Pl. 4/4: Field survey results with an example (architectural ceramics) of the visualisation of the material classes found within individual squares and sub ‑squares. The area of excavation and major finds are marked as well.