Sardis, 2017

Archaeological work at Sardis in 2017 included excavation, conservation of buildings and artifacts, research and publication projects. As always, we are grateful to the Ministry of Culture and Tourism and the General Directorate of Cultural Heritage and Museums for permission to excavate at Sardis, and particularly to Assistant General Director Melik Ayaz, Excavation Department Director Umut Görgülü, and Archaeologist Nihal Metin. The representative of the Ministry of Culture and Tourism this year was Veysel Dağ of the Ephesus Museum. His good will, prudent advice, and consistent help was deeply appreciated by the entire team, and made the season both productive and very enjoyable.

Excavation took place in three areas of the ancient city (figs. 1, 2).

The first, Field 49, is a hill belonging to what we believe was the palace quarter of the Lydian kings (fig. 3). One a goal of our work here over the past 9 years has been to understand the Lydian levels. These are deeply buried under later Hellenistic, Roman, and Byzantine levels, however, and in addition we have found that in most parts of the hill, Lydian buildings have been extensively robbed for spolia. We have evidence for elite occupation on this hill, including pottery, bronze vessels, an ivory furniture inlay, weapons and armor, and fragments of marble building blocks and terracotta revetments; but the buildings themselves were systematically dismantled and reused in later structures, robbing out the walls to a depth of more than 2 meters below ground level.

One of the exciting discoveries of the season, therefore, was a short stretch of a wall built of finely cut limestone ashlars, similar to Lydian terraces, gates, and royal tombs (figs. 4, 5, 6). One ashlar bears a mason’s mark. A burned floor running up to the wall contained Lydian pottery of the mid-sixth century BC, and also three arrowheads, perhaps from the battle between Cyrus the Great and Croesus (fig. 7). On the north, this floor runs up to a much more crudely built wall of schist; despite their very different construction, these two walls belong to the same complex and phase (fig. 8).

We do not fully understand the Lydian architecture here. We had expected to find the continuation of the western Lydian terrace wall, which is preserved about 30 m further south, and whose line is continued in a Hellenistic wall built of Lydian spolia (fig. 3). However, the Lydian walls we discovered are oriented perpendicular to the terrace. Apparently the Lydian terrace was wider here, and the terrace wall must be located further west. We have only uncovered a short stretch of one face, and do not know the thickness of this wall; it is at least 1.7 m thick and may be 4.7 m thick. On the east side of the Hellenistic foundation the Lydian wall were completely robbed to about 2 m below ground level, shortly after the Persian destruction, and only a few stones are preserved. This east-west wall seems to join a north-south wall, which has also been mostly robbed out.

At least three major Hellenistic phases are preserved here, largely composed of reused Lydian blocks (figs. 3, 5, 6, 8). The earliest phase dates to the early third century BC. A peculiar feature is a foundation almost 2 m high, built of very roughly stacked, large limestone blocks, each of which weighs more than 1,000 kg (fig. 9). Beneath this is an earlier phase, also made from reused limestone blocks with anathyrosis and clamp and dowel cuttings. The very coarse masonry of both phases shows that in both periods this was a subterranean foundation; but what did this foundation support?

The north slope of the hill is dominated by a monumental Lydian terrace wall built of large boulders, 3 meters wide and at least 41 m long, and dating perhaps to the 7th or 8th century BC (figs. 10, 11). In 2017 we uncovered a second, even wider terrace wall, also built of massive boulders, more than 5 meters thick, at a slightly different orientation. This seems to date to the late 7th or 6th century BC, and has a face of rather small limestone blocks. A further addition to this structure was exposed to a length of 9.3 m; it bears yet a third orientation (fig. 12). The two walls represent significant expansions and reorientation of the terrace. Excavation within the terrace uncovered more of a mudbrick wall that apparently predating both terrace phases, and articulated bones of a cow or horse within pithos sherds (fig. 13).

Another goal of excavation on this hill was to excavate the Lydian terrace fill to recover dating evidence, and to explore pre-Lydian levels first discovered in 2016. In the center of the hill we excavated through about 6 meters of Lydian terrace fill and learned that much of it dates to the 6th century BC, although the lower strata may be earlier (figs. 14, 15). Beneath this terrace fill, however, were apparently stratified deposits containing large quantities of Bronze Age pottery together with a few small pieces that may belong to the Early Iron Age. Among the Bronze Age forms were at least three bowls with conical “noses,” familiar from Beycesultan and other sites, as well as carinated bowls, grayware and gold-washed wares, and other shapes (figs. 16, 17, 18, 19).¹ Three meters of these deposits were excavated, to a total depth of about 10 meters below surface, without reaching bedrock.

The second excavation area is known as Field 55, and preserves Roman and Late Roman remains (figs. 2, 20). In the early Roman period this artificial terrace belonged to a sanctuary of the Imperial cult; subsequently in the late Roman period it became a relatively prosperous area of housing. Excavation in 2017 was intended to explore late Roman domestic remains (figs. 21, 22). We uncovered one irregular room completely. Its floor was littered with artifacts, probably destruction debris from one of the earthquakes that destroyed this area in the seventh century (figs. 23, 24). Among the artifacts were three hanging lamps, consisting of bronze chains that originally held glass oil lamps (figs. 25, 26); a ceramic flask (fig. 27), a frying pan, an iron horse bit (?), locks, and other objects of iron, bronze, glass, ceramic, bone, and ivory.² A Byzantine lead seal is an unusually late object from this area (figs. 28, 29).³

A massive drain had later been cut through this room and the neighboring room, and led through the terrace wall out to a street beyond. A floor with a round oven, belonging to an earlier phase of this house, was exposed in the cut made by this drain (fig. 30). The chronology and phasing of these houses was clarified through probes below its latest floors.

The third excavation sector is in the northwest part of the city, where a colonnaded avenue entered the Sardis and was spanned by a monumental three-bayed arch (fig. 31). The colonnaded avenue has usually been dated to the fourth or fifth century AD, while the date of the arch is still uncertain but probably in the high imperial period. Excavation showed, however, that the colonnaded avenue is contemporary with the arch, as part of the southwest pier was left unfinished where the south wall of the avenue joined it. This and other indications suggest that the original date of the avenue is significantly earlier than had been supposed.

We completed the area of the south portico, which seems to have had a two-storied colonnade (fig. 32). Doors and a staircase opening to the south under the modern road probably belong to a row of shops similar to the “Byzantine Shops” on the other side of the road. The portico was paved with mosaic, but this was almost entirely missing, probably the result of industrial activity here in late antiquity (figs. 33, 34). A sondage dug against the west face of the monumental arch revealed part of its foundations, and also the foundations of the Lydian retaining wall for the glacis laid against the west face of the fortification.4

Two small buildings are built on the avenue just west of the arch. They seem to be symmetrical, 4 m wide, and entered by a short run of marble stairs on the east. The floors and superstructures were not preserved, however, and the date and nature of these buildings remains uncertain.

A large drain runs along the north side of the street, flanking the north portico of the colonnaded avenue. Although this area had been excavated in the 1960s the mosaic floor of the portico was uncovered for the first time last season (fig. 35). Here at the eastern terminus of the portico, the mosaic bore a dedicatory inscription declaring that “The portico was laid with mosaics under the magnificentissimus count and consularis Flavius Maionios” (fig. 36).⁵ Flavius Maionius, probably a local Lydian citizen whose name is derived from the Homeric name of the region, is known from a verse inscription inscribed on a stone found reused in a later drain that cut this mosaic. This inscription records that “[- - -] Maion[ios] with great [- - -]. But he obeyed and executed skillfully [- - -] a mighty foundation, which is stronger [than - - -].”⁶ The “mighty foundation” might be identified as a major repair to the monumental arch, in which monolithic column drums were packed against its northern foundations (the so-called “Packed Column Area”). Both the repair to the foundations of the arch and the mosaic probably date to the second half of the sixth century.

Site conservation continued to be a major focus of the season. The five-year project to remove biofilm from the temple of Artemis continued, focusing on the columns of the east end. The team of women cleaned seven of the eight columns of the east façade, including one of the two standing columns, transforming this side of the the building (figs. 37, 38, 39, 40). The conservators deliberately left the top two drums of one broken column uncleaned, to document the state of the temple before treatment, and the dramatic change wrought by the project (fig. 41).

A gratifying demonstration of the success of this new technique is that the areas cleaned five years ago have remained white, while areas cleaned recently with other techniques have turned dark again as the biofilm quickly grows back. By using a biocide for an extended period of time - up to a week - we have killed the microorganisms thoroughly without damaging the stone. We expect to re-treat the temple every few years, but this should not be a major undertaking.

At the request of Ministry of Culture, the four displaced capitals from the temple were moved from the locations where Howard Crosby Butler set them up during the excavation of 1910-1914, to a new stoa for protection (fig. 42). In the process we exposed top and bottom surfaces of the capitals which had not been recorded before. All of the capitals are Hellenistic, from the interior colonnade of the building. Two of the capitals (E and F) were found fallen from columns at the east end of the temple. These bear cuttings for their original, Hellenistic dowels and also for Roman dowels from their reuse in the Roman peristyle. They have been extensively repaired after damage during the Roman period. The other two capitals, however, as well as the capital in the Metropolitan Museum, were found far from any standing column, and do not have cuttings for Roman clamps or dowels; they are also far less damaged and do not show wear from rain. After they were removed from the interior colonnade, these three capitals were therefore not reused, but were perhaps displayed as antiquities from the sanctuary, like the collection of ancient sculptures and inscriptions that Butler found set up, apparently in the Roman period, in the northeast corner of the sanctuary.⁷

Research and study projects included Lydian, Hellenistic, and Roman coins, architecture of the temple of Artemis, the Wadi B temple, and other buildings of the city, figural terracottas, Roman sculpture, archaeobotany, conservation of metal artifacts, geology of the region, late Roman buckles, Hellenistic ceramics, and other projects.

• 1Whole or largely preserved vessels include the bowl with conical protrusions (“drum”) P17.059:14805 as well as fragments of similar vessels P17.082 and P17.087:14871:14855; a grayware bowl with lugs or spools P17.063:14812; two plain bowls P17.062:14811 and P17.103:14895. In addition there were a number of small black-on-red sherds probably of Early Iron Age date, which are probably intrusive.
• 2Lamps G17.003:14770, G17.006:14784, M17.029:14785, M17.039:14862, figs. 26, 27; flask P17.026:1473.
• 3M17.013:14715.
• 4The outer face of this retaining wall was exposed in 1990-1991 (Greenewalt, Ratté and Rautman 1994, 13-17, fig. 16.
• 5Now published in Petzl 2019, no. 424.
• 6IN70.007 = Petzl 2019, no. 425.
• 7Suggested already before the capitals were lifted in Cahill and Greenewalt 2016, 503. On the “Nannas Monument” see Butler, Sardis I, 125-127.