Preliminary Report on Excavations of the Archaic Wall

Total Page:16

File Type:pdf, Size:1020Kb

Preliminary Report on Excavations of the Archaic Wall

Carsulae 2008 Preliminary Report on Excavations of the Archaic Wall by Wendy Hallinan

While clearing heavy brush in the area of the fossa at the southern border of Carsulae to begin the first season of the current excavations of the baths, an imposing section of previously unknown archaic wall was exposed (figs. 1). As part of the fifth season of excavations of the baths it was decided to include a preliminary investigation of this closely associated wall. Perhaps one of the most intriguing aspects of this archaic wall is that it is not mentioned in any earlier published works. The maps of the site of Carsulae do not show anything at the southern most area of the town except the baths and the cistern of Imperial date. Umberto Ciotti in an essay on Carsulae actually catalogues the evidence for pre-Roman occupation in the area including similar archaic or “polygonal” walls1 but does not include this wall. In a recent essay Paolo Ciuchini2 studies aerial photographs looking for evidence of defensive walls surrounding Carsulae. He does not mention these large imposing stones. I can only conclude that the overgrowth filling the fossa was so overwhelming that it kept this wall hidden for decades and perhaps centuries. Our experience confirms this hypothesis. While clearing the grass and weeds from the base of the wall, an almost intact oil lamp (08-39, fig. 2) of probable Late Imperial date was found. The lid to a Roman cooking pot appeared in the same general area. A sheet of green glass (fig. 3) of undetermined age was found lodged between the stones of the wall itself. If the area had been previously explored these artifacts would most likely not have been left lying on the surface.

Pre-excavation description of the site The archaic wall runs from southeast to northwest. It is built primarily of large semi- polygonal blocks placed in uneven courses (fig. 4). The limestone blocks are cut to form flat sides but are not all rectangular and vary considerably in both shape and size. The most intact section of wall is over six meters long and two to three meters high. The portion of the wall now visible is up to four courses high and supports a higher area of soil behind it. An additional four meters of wall continues to the northwest, however here it appears that the front or facing blocks of the wall have fallen or been removed revealing a more haphazard arrangement of large limestone rubble behind it (fig. 5). There are several large blocks of stone down below in the defensive fossa. The fossa remains much overgrown and it is impossible at this time to determine the number and size of these stones. Beginning at the northwest edge of the archaic wall a series of stone points appear above ground level running in a line from southwest to northeast - the tops corners of apparently massive stones which seem to form a line at a right angle to the visible section of wall. This line forms a corner with the archaic wall and runs up the slope in alignment with the remains of the Roman cistern (fig. 6). There are six of these stone points spaced two to three meters apart. Bordering the archaic wall at its southeast edge as a three-meter wide gravel road used primarily by goats and sheep. On the far side of this path a short stairway of five marble steps (fig.1) leads up to an area that may be the remnants of the Via Flaminia. This arrangement of archaic wall, possible alignment of stones and the marble stairway form an intriguing group. How are they related? When was the wall built and what was its function? Was it a defensive wall, a retaining wall or a platform? Was there a relationship to the cistern, a possible source of the San Gemini water that fed the bath? Could it have formed an entry to the area of the bath, a transition from the Via Flaminia located somewhere above and east of the baths? Where did the Via Flaminia enter Carsulae from the south? These are some of the questions we contemplated while approaching our first season of excavation. Preliminary attempts to date the wall depend entirely on comparison to similar walls in the immediate area of Carsulae. The two closest examples both in vicinity and construction are in the hill town of Cesi and the walls surrounding the summit of Mt. Erasmo (fig. 7) above it. The dating is unclear and is a current topic of much debate. However Ciotti dates them 5th – 4th century BC and attributes them to Umbrian builders3. Another section of wall in Cesi (fig. 8) dated to the Republican era is more ordered and regular than the archaic wall we are studying. This archaic wall closely resembles the oldest section of city wall of Amelia (fig. 9), which has been dated to the 5th- 4th century BC4. So based on typology alone this wall seems to be pre- Roman from the 5th – 4th century BC. We decided to excavate in two separate areas (fig. 10). The first full quadrant we opened (SB+10 NW) was northwest of the archaic wall between two of the stones that appeared to form an alignment at right angle to the wall. The questions we hoped to answer were: is there another

2 stone between the two visible stones, are they connected to form a wall and where was the ancient ground level? The second full quadrant we opened (SD+11 NE) was at the southeastern end of the wall where it disappeared under the sheep path (fig. 1). We hoped to determine whether the wall continued under the path, if there was a connection with another possible wall coming in from the south at what appeared to be a right angle, and if the stairway on the other side of the path continued its descent along the face of the wall. We were able to answer some of our questions during this first season of excavations. Both quadrants proved to be incredibly productive but in very different ways. In both areas we were stunned by the quantity of artifacts uncovered. In contrast to the internal areas of the baths where primarily architectural features and architectural artifacts remain due to previous excavations, these two quadrants were very rich with ceramic, glass and other small finds. This confirmed our belief that the area had not been previously excavated.

Analysis by quadrant

SB+10 NW (field plan fig. 11, elevation fig. 12, photos figs. 13 – 20) When we opened this quadrant we hoped to determine fairly quickly whether or not the wall turned the corner and continued in a NE direction toward the cistern. We had not anticipated the volume of artifacts we would uncover which greatly slowed our progress. Over 500 pottery sherds (43 drawn) and more than 86 objects were found here. Among the objects were: fragments of 15 oil lamps (10 of them drawn: 08-36, 08-42, 08-43, 08-46, 08-48, 08-50 to 08- 54)), a small bronze cap that may be a miniature helmet or a bell (08-58), 3 bronze coins (one drawn: 08-57) and close to 100 fine glass shards (11 drawn: 08-2, 08-7, 08-8, 08-10, 08-11, 08- 13, 08-15 to 08-18, 08-23). We excavated between two of the rocks visible on the surface looking for evidence of a continuation of the wall. We were able to uncover a 280cm length of archaic wall (field plan, fig. 11). One large rock (C) was found filling the space between rocks A & B. Corners of three additional stones (EFG) were also revealed F & G forming the second, lower, course of rocks (elevation fig. 12). The fully uncovered rocks measure 90-120cm wide, 60-90 cm high and 90- 110cm deep. Their upper surfaces slope down towards the back (SE). One rock (D) was found out of alignment (fig. 13) in the area in front of the wall. It was determined that this rock had

3 most likely fallen from its original location on top of stone C as it was resting on artifact rich soil. Additionally there were several smaller limestone cobbles lying in the space between rock D and rock C (see figs. 10, 13) that are similar to the cobbles filling the area behind the face of the wall. If rock D had fallen one would expect that some of the fill behind it would have fallen after it, landing as these had at the foot of the wall sitting at the same level as the bottom of rock D. The interior of the wall (the SE corner of the quadrant) was filled with limestone cobbles of varying sizes placed rather haphazardly (fig. 14). The artifacts in this area were quite scant. Among the few pieces of pottery were three pieces of vernice nera lying on top of rock C under the overhang of rock B. Some brick and tile fragments were found including one brick still lodged deep in the area of fill. One bronze coin (08-57, fig. 15) was found just above the back of rock C along the eastern balk. A ship’s prow can possibly be seen on one side of the coin and a seated figure on the other. Along the eastern edge of the quadrant what may be the tops of more very large rocks were uncovered. They slope downward from SE to NW with an area of cobble fill between them and the backs of stones A, C and B. This could be the first indication that the wall has an interior face. If so then this is actually a wall and not a platform. Further excavation to the east of SB+10 NW is needed to determine the nature of the wall. Most of the artifacts were in the area along the front of the wall and against the north edge of rock D (fig 16). We thought that perhaps we had uncovered a channel but as rock D proved to be an anomaly there is no evidence for another side to the channel. Approximately 30cm below the surface in the NW of the quadrant we came to a layer of soil with a high charcoal content (fig. 17). This layer continued throughout the excavated area in front of the wall (NW) to a depth of 120cm, which was the full depth of our excavation. There was no sign of burning on the pottery found in great quantity in this layer except for the burning on the exterior of cooking pots. The burned soil and artifacts seem to have been deposited against the face of the rocks perhaps flowing or dumped from an area upslope. The nature of many of the artifacts contributes to a picture of domestic activity occurring nearby. Twenty-one cooking pots were represented among the diagnostic pottery pieces as well as many small bowls, mortaria and 24 amphora fragments. There were several impasto rims of a “pie-crust ware” pattern (fig. 18) that we are hoping to date. Most of the fabric was impasto and ceramica commune although 17 sherds of vernice nera and vernice rossa were found throughout

4 the quadrant including a sherd of a vernice nera bowl at the deepest level of the trench. Of the 15 oil lamp fragments 5 have been identified as vernice nera or vernice rossa. Two interesting and perhaps related finds were a fragment of an oil lamp base (08-48) stamped with “Q.Numice” (fig. 19) and a sherd of a large ceramica comune vessel (08-49) inscribed with “NUM” (fig. 20) Preliminary dating of the ceramics gives us a wide range of dates from one possible pre-Roman bowl and three Republican era cooking pots to several bowls of probable high Imperial Age. Brick and tile were also found here but not in great quantities. Ten examples of the small bricks used in opus spicatum were uncovered. Several fragments of box tiles were found only in the higher loci attesting to our location near the baths. Fragments of roofing tiles, both tegula and cap tile, were found throughout the quadrant. Several shards of fine glass vessels were found, as were two pieces of probable pane glass similar to that found near the baths this season.

SD+11 NE (Field plan fig. 21, photos figs. 22-27, elevation fig. 28) This was a physically difficult area in which to work. The quadrant slopes steeply from east to west. The eastern half is located on the edge of a wide gravel road or sheep path while the western border is approximately 150cm below at the base of the archaic wall (fig. 1). Between the two is an area of loose fall. Visible under the heavy undergrowth just south of the quadrant was a possible additional wall supporting the sheep path above. It seemed to consist of large, unworked, limestone rocks similar to the construction of the interior of the archaic wall. We hoped to determine whether this was another section of archaic wall. Just outside this quadrant was a loose, thin rectangular block of marble that looked much like the steps of the stairway on the eastern side of the sheep path. We hoped to discover its original location. Our most intriguing discovery was made in the eastern half of the quadrant (field plan, fig. 21). In the area where we had thought we might find a continuation of the marble stairway, 50cm below the level of the sheep path we uncovered part of a Roman floor in situ of cocciopesto laid with 45cm square terracotta paving tiles (figs. 22, 23). This floor is aligned with the archaic wall and set against and on top of the rocks of the wall indicating that the floor was of a later building date. The cocciopesto is coved along both the SW and NE edges forming a clear border. More tile is visible along the NE edge. The floor is three tiles or 135cm wide (from SW to NE) and approximately 140cm long (from NW to SE). Along its NW border, the cocciopesto is much degraded however there does seem to be a clear edge above the quadrant’s drop-off to

5 the west. This NW edge continues NE passing beside the front plane of the archaic wall. The SE edge, which is fully paved, disappears into the eastern balk. In the NE corner of the quadrant the terracotta tile continues along the NE edge. The cocciopesto rises along this edge to form a step 20cm above the floor and may step up again another 20cm in the far corner although this area was not fully excavated and it may be the remnants of a wall. A modern pit was cut into the center of the square and continues into the eastern balk (fig. 24) to the depth of the paving. There was a layer of burn along the bottom and sides. It contained much modern garbage including plastic, short metal rods and chunks of cement similar to that used by Ciotti in the baths. This may be evidence of some Ciotti activity in the area. In the NW corner of the quadrant there is a narrow channel formed between the face of the wall and a large rock (rock L) out of alignment with the archaic wall (fig. 25) similar to rock D found in SB+10 NW. The top of rock L is 30cm below the level of the paved floor. We had thought that perhaps rock L was a continuation of the potential wall coming in from the south or the stairway to the east. However no connection to a more southerly wall was found nor did we ever found any convincing evidence that this wall exists. We were able to determine that rock L rests on artifact rich soils and has no firm support so it does not seem to be part of the stairway. There is a space in the archaic wall just SE of the uppermost rock where rock L would seem to fit. The cocciopesto floor begins just SE of this space indicating that it was built when this rock was still in place in the wall. The lower courses of the archaic wall do continue to the SE under the paved floor. The top corner of a lower course of large rocks was also revealed at the bottom of the channel. There were many artifacts found in this quadrant: 125 pottery sherds (17 drawn) and twelve objects (9 drawn). Most of the artifacts were found in the channel discussed above and soil above and in front of rock L. In the area of fall lying on top of rock L the plinth base of a small vernice nera pot (08-55, fig. 26) was removed and is being conserved. We were able to excavate some of the soil lying under rock L; it proved to be rich with artifacts particularly glass and pottery of mixed eras including a fragment of a vernice nera bowl. For the most part the pottery is similar to that found in the other quadrant we investigated (SB+10 NW). It is primarily impasto and ceramica comune and includes fragments of 3 large cooking pots. It differs in that there are many examples here of decorated pottery including two incised pots of vernice rossa,

6 fine sherds of African sigillata and Italian sigillata but none of the “piecrust ware”. Perhaps the artifacts are indicative of an area where food was consumed and not just prepared. The objects include the bottom of an oil lamp (08-41) (fig. 27) stamped “SL” in raised relief and an almost complete (but shattered) oil lamp (08-38). A possible grinding stone of vesiculated basalt (08-59) was also found under rock L, which unfortunately crumbled as it dried out and is now being restored. Four glass shards were drawn (08-4, 08-9, 08-12, 08-14). The smaller sample size is due in part to the more limited excavation at the base of the archaic wall. As we did not open this quadrant until half way through the season, we only excavated here for three weeks. A full investigation at the base of the wall may produce a vast quantity of potentially datable artifacts. Another important find in this area was much painted plaster. These fresco fragments were found throughout the quadrant and formed a visible stratum 10cm thick resting on top of the tiled floor (fig. 28). Many were of an ochre color with some lines of blue and red. The quantity of plaster and its location directly in contact with the floor suggests that the walls of this tiled space were plastered and frescoed. We did not find a continuation of the stairway from the other side of the sheep path. It may be that we did not excavate deeply enough and it is buried beneath the cocciopesto floor. There is a layer of fine yellow sand just under the cocciopesto, which can be an indication of rebuilding in Roman sites. Another hypothesis is that the floor was part of the stairway – perhaps a portico or covered landing. Further excavation should answer some of these questions.

Conclusions

Our first year of excavation along the archaic wall produced many perplexing surprises but also answers to some of our preliminary questions. We now know that the wall turns in a NE direction towards the Imperial aged cistern. We also know that it continues to the SE under the sheep path and does not connect with a wall coming in from the SW. So it forms a corner of a defensive wall or two sides of a retaining wall or a platform. Its function is unclear as is its relationship to the cistern, the baths, the stairway and the Via Flaminia. Our big surprise was that a Roman floor was built abutting an upper edge of the archaic wall but we do not understand the nature of this floor. It is narrow (135cm) and its current

7 exposed length is 140cm but it clearly continues under the sheep path. We did not reach the ancient ground level at the base of the wall in either of our areas of excavation and never produced any datable stratigraphy. The artifact and charcoal rich layer uncovered in SB+10 NW is over a meter deep with no discernable change in the type of finds with dates ranging from Republican to Late Imperial. This area bears some resemblance to what are called basis villae. These sites are characterized by a platform built of polygonal stones and many ceramic finds of coarse ware pottery. However the location within the city of Carsulae makes it an unlikely site for a rustic Republican villa. More investigation is needed to totally rule out this hypothesis. This type of wall has proven very difficult to date throughout Umbria and the date of this archaic wall is certainly not clear. The fact that vernice nera sherds were found in among the large polygonal stones suggests a 4th - 3rd century BC date. However further excavations both along the foot of the wall and its interior may provide more exact dating information. We cannot say with any certainty whether it is Umbrian or Republican. Next season I suggest that we also open a quadrant just south of the cistern where the newly uncovered section of the archaic wall seems to meet it both to determine if there is continuity between them and to see if there is any evidence of an ancient water source. We should also open the quadrant SD+12 NW to determine the extent of the paved floor and the floor’s relationship to the stairway. I would hope to have the opportunity to study this area further as it has the potential to provide the information needed for us to gain important insights not just into the origins of Carsulae but also into the early era of Romanization of Umbria.

Respectfully submitted,

Wendy Hallinan

8 9 1 Ciotti, U. Carsulae e San Gemini p 16-17 2 Ciuchini, P. “Carsulae, Alcune Notazioni…” in Archeologia Aerea 3 Ciotti San Gemini e Carsulae p. 16 4 Bradley, G. Ancient Umbria, State, Culture, and Identity in Central Italy from the Iron Age to the Augustan Era p. 77-80

Recommended publications