The study of the Roman lamps and pottery from the Sanctuary of Demeter and Kore is based on the finds from the excavations of 1961 to 1975 on the slopes of Acrocorinth (Plan 1). The inventoried and context material of Roman date comes from the whole of the area of the earlier Greek Sanctuary (Plan 2), but the greatest amount of Roman finds and the better stratified material are limited to the Upper Terrace, the eastern half of the Middle Terrace, and the area in and around Building K-L:21-22 on the Lower Terrace (Plan 3); elsewhere within the earlier Sanctuary, significant Roman material came only from the Early Roman dumped fills over Buildings M: 16-17 and L-M:28.
The sloping topography of the site created several difficulties for the excavation and study of the material, particularly in the upper levels. The surface of the site was covered by a sloping layer 0.15 m. or more thick, in which the material was fragmentary and very battered and worn; the material within this surface layer was often mixed in date, spanning the whole Greek and Roman occupation of the Sanctuary and occasionally including pieces of 5th to 6th century, Byzantine, Turkish, or modern date. The layer appears to represent wash down the slope of the hill in the centuries intervening since the abandonment of the site.1 The incongruity between this sloping surface layer and the horizontal occupation levels below it offered the constant possibility of contamination of the lower levels. It also appears, from the fact that strata of considerably earlier date (sometimes even of the Greek period) often immediately underlie this layer, that parts of the stratigraphic sequence are missing, presumably washed farther down the slope. This condition will partly explain the present predominance of earlier Roman material on the site.
The effects of this slope erosion are reflected as well in the poor state of preservation of the pottery. Much of the wear on the edges probably occurred as the sherds rubbed against each other and the surrounding soil as they were washed downslope. But in addition, the bulk of the Roman ceramics from the Sanctuary has a soft, powdery surface, alien to its original form, and frequently a lighter color;2 when found, much of the material was encrusted with a tough, calcareous deposit. It seems clearly indicated that these effects are a result of surface and subsurface water run-off on the slope. The condition was least pronounced in the pottery from the burnt-grain packing behind Building T-U:19 (lot 73-97) and on the floor of the same building (lot 73-98), where the pottery was protected by a high bedrock scarp on the uphill side of the building; hence it was not so exposed to run-off water or the consequent and repeated alternate wetting and drying out. The condition greatly hindered the identification and differentiation of fabrics, particularly the unknown coarse and cooking fabrics,3 and the attribution of these pieces should be regarded as provisional.
The excavations of the Demeter Sanctuary produced so much material that storage was a serious problem. Throughout the excavation it was customary, after preliminary washing, to discard body fragments of coarse and cooking pottery and sometimes of fine pottery as well. Then in 1970 the storage problem became acute. The finds from the previous years of excavation, particularly those which appeared very mixed in date, i.e. the surface layer, were once more sorted, "thinned", and in some cases combined. In later years small groups of material and those not significantly associated with architecture were normally discarded after being preliminarily described, and body sherds, particularly of coarse and cooking wares, were discarded after description and mending even from such closed groups as lot 73-97. These practices have once again affected the later Roman material more than the early Roman, and the coarse and cooking pottery more than the
1 This opinion was confirmed by the distribution of a few vessels and pottery types. For instance the phialai Corinth XVIII, i, nos. 675-681, pp. 191-194, were found primarily on the slope below Building T-U:19 and over Building O-P: 19-20. A second instance is possibly the fragments of unglazed Corinthian lamps found upslope and downslope of the north wall of "Building" P-Q: 25-26; for this group see pp. 15-16 below.
2 The most striking example of this is 66, but it is true for almost all the pottery and lamps.
3 The alteration appears in most cases to have proceeded completely through the biscuit, so that even breaking sherds to secure a freshly broken surface could not solve the problem.
lamps and fine wares. Also as a result, it is not possible to give statistics for the totality of finds from the excavation. A careful count of all the Roman fragments still preserved, however, does indicate that similar contexts, such as the various parts of the surface layer, tend to have a similar chronological mixture of finds, and that single categories of finds, particularly those of limited chronological range such as unglazed Attic lamps or Italian sigillata, are repeatedly found in similar amounts in contexts of similar date.4 It therefore appears that what remains of the Roman material from the Sanctuary closely reflects what was found, especially for the lamps and fine wares. In the discussion of each category of object, I have given the number of pieces found, and in the discussions of the lamps and of the dates of the lots I have indicated percentages.
In studying the Roman ceramic material from the Sanctuary I have been concerned with it primarily as an archaeological tool, to be used as evidence in the discussions of the stratigraphy and history of the Sanctuary and of the cult practices, which will appear in subsequent parts of Corinth XVIII. I have directed my attention, therefore, to identifying the material and its sources, to establishing its chronology and the chronology of its contexts, and to determining, as far as possible, its function. The last is of more than ordinary importance for the Demeter Sanctuary, first of all because the outlying and relatively isolated location of the site suggests that the material was used more or less where it was found,5 and second because the purpose of the site is known and the function of the ceramic material will therefore throw light on the cult. With these aims in mind, the scope of the study was broadened to include not only the Roman pottery but also the Roman lamps from the Sanctuary.
The selection of material for the published catalogue was controlled by several factors. The primary consideration has been to present a corpus of Roman material representing the range of common and unique finds from the Sanctuary, with aims identical to those described above. Secondly, although the repertoire in the Demeter Sanctuary is more limited than that from other parts of the Corinth excavations, it corresponds in kind (excepting only a few categories). Since no previous report on Roman pottery at Corinth in extenso has appeared, this study will serve as an introduction to the pottery of the first four centuries after Christ, both imported and local. The lamps and the imported fine-wares, which can be dated by parallels from Corinth and elsewhere, on the whole reflect the history of the Roman Sanctuary. I have attempted to give an outline of the typology and history of the coarser wares, which are primarily of local origin. The results must be regarded as preliminary until they have been tested by systematic study of the whole body of Corinthian pottery, and by further excavation.
Because the lamps are easily identified even in small fragments and because they were often retained when pottery was discarded by the excavators, the study of the lamps provided the chronological basis for the pottery study; the Roman lamps have therefore been presented first. This framework was then checked and modified as necessary, as the study of the pottery progressed. Recent pottery studies have favored arrangement of material by deposits in order to facilitate chronological and particularly quantified studies. Such an arrangement was inapplicable to the Demeter pottery, in large part because the mixed range of many of the strata and the paucity of closed deposits would have eliminated much of the material. In addition, since much pottery had been discarded owing to lack of storage space, no group remained intact at the beginning of this study. The pottery is therefore arranged by ware and function. It begins with the fine table wares, chronologically presented, as they provide some additional dating evidence as well as a quick synopsis of the
4 That is, the range of Italian sigillata, for example, is basically similar in all contexts of the third quarter of the 1st century or in all contexts of the first half of the 2nd century, and while naturally smaller in amount, the ware shows the same chronological and typological range even when it is found in surface layers.
5 This circumstance is relatively rare in Roman Corinth, where the excavators' attention mostly has been directed to the Roman forum and its vicinity, and the Roman pottery mainly comes from large secondary fills associated with the construction or remodeling of public buildings. The location of the Demeter Sanctuary suggests, however, that even in the case of the secondary fills which form most of the Roman strata, the material was used within the excavated area.
chronological limits of the Roman use of the Sanctuary. They are followed by the "ritual vessels", which are of both fine and cooking fabrics, imported and local, and which are important for the study of the Sanctuary and the cult. The next sections deal with the related imported and local cooking fabrics, first the cooking-pots themselves and then the closed forms not for use with heat (thin-walled wares and pitchers). Then the amphoras and miscellaneous coarse wares are presented. In the last section, the grave gifts from the cemetery, which in Late Roman times succeeded the Sanctuary on the site, are discussed.
As elsewhere in the Corinth excavations, the number and variety of imports in the Sanctuary are greatest in the 1st century, when not only fine wares from Italy, Gaul, Asia Minor, and the Black Sea and western and eastern amphoras were imported but also thin-walled wares and cooking wares (possibly including a thymiaterion) from Italy, pitchers, and even coarse pottery (probably from the northern Peloponnesos). This rich variety falls off slightly in the 2nd century. From the late 2nd century through the 4th century the imports are basically confined to Attic lamps, fine wares from Pergamon and North Africa, and amphoras, with a few Italian thin-walled pieces and cooking pots of unknown origin imported in the 3rd century.
Throughout this period, the imported pottery must have come by sea from west and east. Little can have traveled to Corinth by road, as no Attic (or Argive) pottery of the 1st and 2nd centuries is found and only a few Attic pieces of 3rd- or 4th-century date can be recognized in the Sanctuary. The lamps are in striking contrast, since they imitated Italian but not eastern types in the 1st century; moreover, the only lamps imported in quantity were the Attic lamps of the mid-3rd to second half of the 4th century. The quantity of imports from a single source, the rather lengthy periods over which some of the wares are represented, and the lack of congruence between the sources of lamps and pottery, not to mention amphoras, strongly suggest that we are dealing here with deliberate and extended trade in luxury goods6 rather than single shipments or chance acquisitions by travelers. As I have tried to show in the discussions and in the catalogues, the imports found in the Sanctuary were readily available on the Corinthian market. Only the imported thymiaterion 148 and some of the special coarse pottery from the burnt-grain packing behind Building T-U:19 are likely to be special imports.
In general the fabrics have been described under the individual wares or, in the case of rare or doubtful fabrics, under the catalogue entries. Attic fabric is dealt with extensively in the discussion of unglazed Attic lamps of the 3rd and 4th centuries and more briefly in the sections on Attic glazed lamps and Attic post-glazing lamps (pp. 17-18, 19, 21 below). Various Corinthian fabrics are also described at several points in the discussion,7 but it seems worthwhile to summarize those descriptions here.
The Corinthian fabric ordinarily employed for Roman fine wares is buff, soft to medium in hardness, and gritty in texture. Its color is normally reddish yellow (5YR-7.5YR 7/6), sometimes very pale brown (10YR 8/3), depending on the kiln atmosphere; glazes, where they occur, are slightly lustrous red (2.5YR 5/6) or matt black, respectively. Except for the unglazed Corinthian lamps, which have a very smooth texture, this fabric is normally gritty with small dark and white inclusions and some voids. The coarser grades, which I arbitrarily term "Corinthian coarse", are not so gritty, having medium-size rather than small dark and white inclusions, fewer voids, and medium to large red inclusions (sometimes grog or hornfels) and sand; they are normally thicker walled than the finer grades. This fabric, in its various grades of coarseness, is used for lamps, imitations of sigillata and other fine wares, and domestic coarse wares. The second major Corinthian fabric I have termed "Corinthian cooking", as it is the fabric used for local cooking pots. Lighter in weight, more brittle, harder, and more heat absorbent than the Corinthian buff fabric, the Corinthian cooking fabric is normally red or reddish brown (2.5YR 5/6—5YR 5/4) with angular, non-calcareous white inclusions and fewer weathered orange and dark inclusions; mica is occasionally seen. The exterior surface or the whole
61 use the term here rather loosely to refer to products, such as lamps, fine pottery, and cooking pots, which were traded for their own qualities rather than for their contents, as amphoras were.
7 The reader is particularly referred to the discussions of Corinthian glazed lamps, unglazed Corinthian imitations of Attic lamps, local cooking pots, and miscellaneous coarse and cooking wares (esp. pp. 12, 21, 74, 119 below).
vessel is not infrequently fired gray. This fabric, too, is found in various grades of coarseness, from very fine to medium coarse depending on the product. The thickness varies, too, but vessels of this fabric are thinner walled, size for size, than vessels of the buff fabric. In increasing grades of coarseness the products made of Corinthian cooking fabric are wheelmade lamps, thin-walled wares, and miscellaneous small shapes such as bowls, cooking pots, four series of pitchers, thymiateria with piecrust decoration, and pedestal craters, and probably a local series of amphoras.
The overwhelming majority of the Roman lamps and pottery from the Sanctuary of Demeter can be dated between the middle of the 1st century after Christ and the third quarter of the 4th. That there is a preponderance of material of the second half, especially the third quarter, of the 1st century is due to the construction fills and contemporary material associated with the three buildings on the Upper Terrace and Building K-L.21-22 on the Lower Terrace. Concentrations of material of the first half of the 2nd century were found southeast of Building K-L.21-22 (lots 73-103, 73-104), and various fills of 3rd-century date were recovered at the east end of the Middle Terrace. Lamps, pottery, and coins of the third quarter of the 4th century were ubiquitous in the destruction debris of the buildings and in the surface layer.
Of particular interest is the large quantity of well-preserved pottery of the second and third quarters of the 1st century found with large quantities of burnt grain in a packing between the walls and the bedrock cutting for Building T-U:19 (lot 73-97); joining fragments for the material in this packing were found in the robbing trenches and destruction debris over the building and also to its east, where the packing had slumped or been disturbed in later times. This is the earliest coherent Roman deposit in the Sanctuary, and it is unfortunate that its deposition cannot be securely connected to the construction of the building. Particularly useful for establishing the date of the packing, which unfortunately contained no coins, were the Corinthian wheelmade lamps 1 and 2 of the middle of the 1st century and the moldmade lamps 12 and 14 which could be slightly earlier, the Arretine plate of Camurius 80, and the imported trefoil oinochoe 214 and pie-fluted jar 270. The Italian sigillata cup 77, the eastern sigillata A plate 90, and Pontic sigillata plate 101 from lot 2154 to the east of Building T-U:19 and the eastern sigillata B plate of f Epfxr/s 92, the partly glazed jars 138-140, and the Italian thin-walled cup 194 should probably be associated with the deposit.
The lower filling of Cistern 64-1 on the Middle Terrace (lot 2100) has received wide notice as a deposit of Herulian date.8 Scattered through the fill were 17 bronze coins, of which one dissolved in cleaning; the latest were seven of the time of Gallienus (a.D. 260-268) and one of Valerian struck in 257.9 In addition to the catalogued pottery, of which the two African red-slip plates of Hayes form 50A (110 and 111) are surely mid-3rd century and the Aegean amphora 254 is 3rd century, 124 sherds of the mid-2nd to the second half of the 3rd century remain in the lot. Most diagnostic are the lamps: one Corinthian wheelmade lamp, one glazed fragment of lst-century type, two unglazed Corinthian lamps with vine pattern on the rim and rays on the discus,10 one unglazed Attic base with a central boss and three framing circles, an unglazed Attic lamp paralleled in the 1982 deposit mentioned above (50), and another of the Leaf Shop with goddess with double ax on the discus and 8-S pattern on the rim (42).
-+ Hesperia 37, 1968, pp. 309-310; Kenchreai V, p. 40, note 44 (incorrectly reporting the lamps). Synkellos, p. 717 (W. Dindorf, ed., Bonn 1829) writes that the Heruli attacked Athens, Corinth, Sparta, and Argos, and the whole of Achaia, but destruction deposits have been published so far only from Athens (see particularly Athenian Agora VII, pp. 20, 62-63 for references) and Olympia (A. Mallwitz, Olympia und seine Bauten, Munich 1972, p. 112 and E. Kunze, "Zur Geschichte und zu den Denkmälern Olym-pias," in 100 Jahre deutsche Ausgrabung in Olympia, Munich 1972, pp. 24-25). Other material of this period was found in the East Field beside the Sanctuary of Poseidon at Isthmia (Jeanne Marty, personal communication). At Corinth itself, the "Herulian" debris in the South Stoa, identified by Broneer on the basis of a hoard of appropriate date (Corinth I, iv, pp. 135-138,159), contains Attic glazed lamps of the 4th century. Another hoard of this date was found in the Shear excavations in the Theal -+ [Shear, A]A 35, 1931, esp. pp. 146-151) without associated pottery. For a deposit of ca. 250 from the area east of the Theater, -+ Williams and Zervos, Hesperia 52, 1983, pp. 14-18.
9 For a complete list, see p. 136 below, lot 2100.
10 For the date of this subgroup, see p. 16 below.
Three factors create difficulties: 1) the fill of the tunnels which open off the lower part of this fill (lots 2108, 2109) was no later than 2nd century, and it therefore appears that a second, earlier fill may be included in lot 2100; 2) Perlzweig states specifically that the type of goddess with double ax "did not appear on Attic lamps until after the Herulian invasion, a.d. 267";11 and 3) no other debris of mid-3rd-century date was found in the Sanctuary, as one would have expected if the destruction were caused by the Heruli. One might get around the second difficulty by suggesting that 42 actually represents contamination from the 4th-century filling above (lot 2099), but the lamp was mended from several fragments, all of which were in lot 2100, and this explanation is therefore unlikely. Thus it seems probable that lot 2100 is a filling dumped into Cistern 64-1 in the late 3rd/early 4th century and that it cannot be associated with the Heruli. I have therefore assigned the second half of the 3rd century as a general date for the lot.
No destruction debris from the Sanctuary was apparently found in situ. For a discussion of the date and contents of the debris associated with Building T-U:19, see p. 20 below.
The question of the date of the earliest and latest material from the Sanctuary remains to be discussed. As regards the early material, I have already discussed the earliest deposit from the Sanctuary, the burnt-grain packing behind Building T-U:19 (lot 73-97), which contained material of the second and third quarters of the 1st century. A number of other strata were also datable to the third quarter of the century,12 and the continuous sequence of Roman lamps and pottery seems to start at that time. One notes in particular the lack of early examples of Corinthian wheelmade lamps, of late Augustan Italian sigillata, of eastern sigillata A in general, and of imported thin-walled wares such as appeared in the Tiberian deposit in the forum.13 A few pieces must be dated earlier than the middle of the 1st century, however. The two Ephesos-type lamps (11 and a second piece in the same lot), the Italian sigillata plate of Haltern 1 (73) and two contemporary fragments, perhaps the eastern sigillata A cup imitating Arretine Service I (91, although this could be second quarter of the century), a single fragment of eastern sigillata B, and the dot-festoon-decorated beaker 192 are all the pieces that definitely can be dated early. Fragments of the thymiateria with pie-fluted decoration appear in contexts with little or no other Roman pottery (e.g. 145-147) and conceivably could be dated early as well, but other pieces, 149 and 151, of which the first is imported, are found in contexts of the third quarter of the century, and parallels in the West suggest this date or later is an appropriate date for all. Of the early pieces listed above, all probably date after the founding of the Roman colony at Corinth in 44 b.c. 192 is the earliest, 11 may be next, or it may be middle to late Augustan with 73 and its companions and possibly 91. This quantity of material seems insufficient to demonstrate that the Sanctuary was in use before the middle of the 1st century. It probably represents pieces casually dropped in the area by passers-by, with an increase in activity in the latter part of the period.
It remains to discuss the latest material from the Sanctuary. The lamps sharply decrease in quantity late in the production of Attic glazed lamps, just as the post-glazing lamps were beginning. In the section on lamps (pp. 20-23 below), I have argued on the basis of context for a date in the third quarter of the 4th century. Here I may add that the sequence of fine wares ends with African red-slip-ware forms (111-115, and see p. 55), which are now dated to the third quarter of the 4th century on evidence quite removed from Greece and different from the chronological evidence for Attic lamps. A terminal date for the bulk of the Roman finds14 from the Sanctuary in the third quarter of the 4th century may therefore be established.15
11 Athenian Agora VII, p. 117.
12 The material of this date is readily identifiable at Corinth, because there are a number of large groups in the forum which represent clean-up of the debris of a Vespasianic earthquake, probably in a.d. 77. For the earthquake, see Corinth VIII, ii, no. 20, pp. 18-19, with references to Johannes Malalas and Plutarch; for an alternative date for P. Licinius Priscus Juventianus, see now Corinth VIII, iii, no. 306, esp. p. 121. The groups publishec -+ Hayes, Hesperia 42, 1973, probably represent such clean-up. For another similar group and a discussion of the date of the earthquake, see my article, Slane, Hesperia 55, 1986, pp. 315-317.
14 Aside from the fine ware and lamps, the ceramic material from the Sanctuary cannot be precisely dated except by its presence in the Sanctuary. For these pieces I have preferred to give a more general terminal date, second half of the 4th century.
15 Two destructive earthquakes shook Corinth in the third quarter of the 4th century, in a.d. 365 and again in a.d. 375. The sources have been collectec Wiseman, Hesperia 36, 1967, p. 409, note 19. The chronological coincidence is appealing, but conclusions must await the completion of the other studies of the Demeter Sanctuary.
Once again a few pieces are apparently of slightly later date, lamp 58, a second Attic post-glazing lamp in lot 4368, and dish 120 in particular. Of these I am inclined still to place 58 in the third quarter of the century on the basis of its parallel with the shop T-. All three pieces differ from the bulk of 4th-century material in being nearly complete. As the lamp in lot 4368 and 120 can be comfortably dated about the middle of the 5th century, it is tempting to associate them with the Early Christian cemetery on the site. From the point of view of the ceramic material, the site of the Sanctuary seems to have been abandoned from the third quarter of the 4th century until about the middle of the 5th, when the cemetery was established. 120 can probably be associated with 276 as a grave gift, the lamp in lot 4368 with 62 and its companions either as grave gifts or as pieces casually dropped by visitors to the cemetery.
The function of particular classes of ceramics having been discussed in the appropriate sections (see, in particular, lamps, ritual vessels, thin-walled wares, and amphoras), the general conclusions that can be drawn concerning the function of the material in the Sanctuary remain to be considered. A few classes are clearly votives or cult furniture (the over-size lamps and the thymiateria), but the majority of the material differs not at all from what would be found in any domestic or secular context. Some vessels are suitable for pouring (the unguentaria, some thin-walled wares, and the pitchers) and might have been used for libations. The association of many thin-walled wares and Corinthian wheelmade lamps in the early phase of Building K-L:21-22 on the Lower Terrace might suggest libations and lamp-lighting as rituals at the entrance of the Sanctuary, but even if so, the practice appears to have been discontinued after the 1st century. The cooking pots, which are mostly large, have all been used, suggesting that cooking took place in the Roman Sanctuary, and thus the fine-ware plates and the less numerous cups are probably just what they appear to be, vessels for eating meals prepared in the Sanctuary and for drinking.
Was this article helpful?