The Corinth Oinochoe: One- and Two-Handled Jugs in Ancient Corinth

Abstract

One of the many characteristic shapes produced in Corinth during the Archaic and Classical periods was the round-mouthed jug with one or two handles, the so-called "Corinth oinochoe." The present article examines the typological development of this shape, particularly the version with two handles, from its introduction in the late seventh century until 146 B.C. This development suggests changing customs in male dining at Corinth, particularly in the third quarter of the fifth century and at the end of the fourth. The function of the two-handled variety is briefly considered, as well as the evidence for ceramic connections between Corinth and Athens.

Introduction

In 1971, Charles K. Williams II, then Director of the Corinth Excavations, was working in the pre-Roman levels in the area of the Forum at Corinth, uncovering a group of large, if at times poorly preserved, buildings of the fifth and fourth centuries just to the north of, and partially under, the colonnade of the South Stoa erected in the Early Hellenistic period.¹ Between Buildings I and II, Williams came upon a drain (1971-1), really a round-bottomed water channel, cut into the soft bedrock and sloping down from southeast to northwest toward the major stone drain that ran along the north side of Buildings I–III.² At some time in the later fourth century, when Buildings I and II went out of use, the area was leveled and the drain was filled with a large accumulation of broken pottery, figurines, and other objects.³ Since the quantity of pottery from the drain is large and all was kept, it can be studied both typologically and statistically, and will provide a cross section of the local and imported shapes in use in this area of the city during the fourth century, particularly the latter half, when the city seems to have undergone significant social and economic change.

Figure 1.

Two-handled Corinthian "decanter" or "mushroom jug": Edwards's Decanters I, II, and III. After Corinth VII.3, pp. 58, 60, 62, nos. 283, 287, 302, pls. 11, 12. Scale 1:3

Among the pottery from the drain was a type of round-mouthed jug furnished with two handles, normally set about 40–50° apart, a "decanter" or "mushroom jug," as it has been termed in the archaeological literature, one of the more distinctive shapes produced by the potters of ancient Corinth during the fourth century.⁴ The principal scholarly discussion of the two-handled jug at Corinth occurs in G. Roger Edwards's fundamental volume, published in 1975, on the Hellenistic pottery from the city.⁵ On the evidence then available, Edwards believed that the form was first developed "in the later Archaic period" and that its production was "sporadic." Excluding three "short-lived experiments" (C-1937-2056, C-1939-22, C-1936-1116) in the period ca. 550–450/425, Edwards felt that the Corinthian potters evolved three main types, which he termed Decanters I, II, and III, respectively (Fig. 1). Decanter I was represented by four examples from a single deposit of 460–420 and one possible Hellenistic successor, perhaps of the first half of the second century B.C. Decanter II was also developed in the third quarter of the fifth century, the earliest examples being placed, on the basis of context and of the theoretical development of the shape, "well within the third quarter" of the fifth century, the latest in the early first quarter of the fourth century, although Edwards suggested that the type might have continued into the second quarter. Decanter III was a "modified continuation" of II. The earliest preserved examples belonged to the third quarter of the fourth century, but production might have begun in the second quarter, so that types II and III might have briefly overlapped. Decanter III enjoyed a short life, the last examples dating to the first quarter of the third century. These three types thus covered the period from the third quarter of the fifth perhaps to the first half of the second century B.C.

A study of examples from excavations of the last forty years at Corinth, combined with a reevaluation of the earlier material, has made it possible to recognize more clearly the relationship of the two-handled jug and the one-handled Corinth oinochoe of the Archaic period, and to trace the gradual evolution of both versions of this jug from the late seventh century to 146 B.C., and perhaps even later. This typological study forms the main part of the present article. I also briefly discuss the functions of this type of vessel and the evidence provided by this shape for connections between potters in Corinth and Athens, especially in the second half of the fifth and the first half of the fourth century.

The One-Handled Corinth Oinochoe

During the last quarter of the seventh century (perhaps ca. 620), at the beginning of the Early Corinthian (EC) period, potters at Corinth developed, initially for local use, a distinctive form of jug with a bulbous body, broad neck, and round mouth, and with one or, as we shall see, two handles rising above the level of the rim. Patricia Lawrence, in her publication of the Anaploga Well (well 1962-5), named this shape the "Corinth oinochoe" and gave an excellent account of the development of the type with single handle to ca. 560, near the end of Late Corinthian I.⁶ What prompted the introduction of a new type of jug is unknown, but the shape seems to combine a squat version of the bulbous body of an olpe with the high broad handle that was long used by Corinthian potters for jugs.⁷

The earliest stage in the development of the shape is represented by such vases as C-1962-569, C-1932-134 (Fig. 2), and C-1940-87, all of which belong to Early Corinthian (620/615–595/590).⁸ In this first phase, the vase is generally not large, the height in the three examples cited ranging from 16.4 to 17.1 cm, the diameter of the body from 16.3 to 17 cm. The base is broad, 10–11.3 cm in diameter, with a low, well-tooled ring-foot. The body is baggy, with a low point of greatest diameter. The neck is short, about 2.2–2.5 cm high, and is everted. The rim tends to be flat or only slightly rounded on top. The handle of C-1932-134 takes the form of an elongated oval, about 2.9 cm wide, and rises high above the rim.⁹ Early examples of the shape may carry figured decoration (including "padded dancers," as in Fig. 2), but almost from its inception, a second variety was produced, which was semiglazed, usually with one or more broad, horizontal bands on the body.

By Late Corinthian I (ca. 570–550), a canonical form had been evolved, of which C-1962-395 (Fig. 3) may serve as an example. The size is quite standard, usually about 17–18.5 cm in height. The shape exhibits a low, flaring ring-foot; a body that is now more globular; a taller neck (ca. 3.3–4.1 cm high) with vertical or slightly flaring profile and often separated from the shoulder by a ridge (or drip-ring, as in other shapes, e.g., the olpe); a flat rim (ca. 0.8–1.2 cm wide); and a broad strap handle (normally ca. 2.9–3.2 cm wide) that still loops high above the rim.¹⁰ As Lawrence has shown, figured decoration was uncommon after the beginning of Late Corinthian I and had ceased by the end of the period. By this time, the one-handled Corinth oinochoe had become a light-colored vase articulated with only a few black bands—usually a single band above and below the point of greatest diameter, a third on the edge of the foot, and a fourth on the top of the rim. The development of a utilitarian shape decorated in a relatively simple manner is perhaps typical of conservative Corinthian taste in pottery for home use.¹¹

Figure 2.

One-handled oinochoe C-1932-134 from well 1932-4. Scale 1:3

For the development of the shape during the next century, from the middle of the sixth to the middle of the fifth century, we are primarily dependent upon four deposits at Corinth: wells 1937-3, 1947-1, and 1939-1, and pit 1975-1.¹² At least eight one-handled oinochoai were found in well 1937-3, the pottery of which covers the second half of the sixth century and the beginning of the fifth. There is again a variation in form, which may be indicative of chronological difference, though this cannot be definitely affirmed.¹³ The two very fragmentary jugs C-1937-2470 (Fig. 4, top left) and C-1937-1060a–c seem closest to the latest Anaploga examples. The first preserves a broad rim, glazed on top, a short neck, and a ridge at the base of the neck; it may have had a globular body, judging by the curve of the shoulder.¹⁴ The second also has a ridge, and preserves a globular body, with two broad bands, and the old ring-foot.¹⁵ However, with C-1937-997 (Fig. 4, top right) we see the introduction of certain new elements. Although 997 retains the well-defined ridge articulating the separation of neck from shoulder, and probably had a neck profile similar to 2470, the body now has a somewhat higher point of greatest diameter, effectively inverting the old body shape, and comes closer to the body of a trefoil oinochoe. In addition, C-1937-997 has a disk-foot, with slightly concave underside. More innovative is C-1937-1017 (Fig. 4, center left), for, although it retains the broad rim (probably glazed) and the two broad bands on the body, it has the new disk-foot, and the globular body passes without a break into the short, outcurving neck.¹⁶ Finally, C-1937-2469 (Fig. 4, center right) seems to represent an even more radical transformation of the older form: it has an almost horizontal shoulder with slight jog at the base of the neck; a broad, strongly flaring neck; a thickened rim; a strap handle concave in section; and a disk-foot. The potting is very fine, the surface highly polished; and the system of narrower red and brown-black bands, particularly the three red bands on the inside of the neck, is, for this shape, quite new. It would not be surprising if 2469 were the latest of the Corinth oinochoai from this deposit, datable to the late sixth or very early fifth century.¹⁷

Figure 3.

One-handled oinochoe C-1962-395 from well 1962-5. Scale 1:3

Comparable to C-1937-997 are C-1947-163 (Fig. 4, bottom) and C-1947-164, both from an unfinished well or pit along the east side of the Southeast Building (well 1947-1), the contents of which are roughly contemporary with those from well 1937-3. Fabric, shape, and decoration suggest that the two pots were made in one workshop, perhaps active in the late sixth or early fifth century. C-1947-163 preserves a flat disk-foot with beveled edge. Both 163 and 164 have a body that is rather more ovoid than globular (decorated with two broad bands); a distinct ridge at the base of the neck; a tall, slightly flaring neck, ending in a broad, flat rim painted red on top; and a typical strap handle.

A more prolific deposit, well 1939-1, with pottery ranging from the second quarter of the sixth century to the early third quarter of the fifth, has produced at least eight inventoried examples of the one-handled Corinth oinochoe, C-1939-113 to 120. These have been studied by Julie Bentz

Figure 4.

One-handled oinochoai from various contexts. Scale 1:3

in an unpublished dissertation.¹⁸ On the basis of shape, she has arranged the vases in the following chronological order:

• 1 C-1939-117, 118 third quarter of the sixth century, perhaps into the fourth quarter

• 2 C-1939-119 fourth quarter of the sixth century

• 3 C-1939-115, 120 first quarter of the fifth century

• 4 C-1939-113, 114 second quarter of the fifth century

• 5 C-1939-116 mid-fifth century

Figure 5.

One-handled oinochoai from well 1939-1. Scale 1:3

Although this arrangement is probably broadly correct, some of the distinctions are perhaps overly precise, and some allowance may need to be made for differences between workshops and between the mannerisms of individual potters. Certainly C-1939-117 (Fig. 5, left), 118 (Fig. 5, right), and 119 are likely to belong to the second half of the sixth century. All three have the broad bands on the body that are characteristic of this shape in the sixth century, but the profiles differ, so that they may represent the work of three different potters or workshops. C-1939-117 has the same unusual form (and banding) as C-1937-1017: they may even come from one workshop. C-1939-118 and 119 are more normal, with ring-foot, groove separating shoulder from neck, slightly flaring neck, and horizontal rim painted red. C-1939-118 has an ovoid body but that of 119 is considerably squatter, rather like C-1937-997. The remaining vases from well 1939-1 probably date to the first half of the fifth century. C-1939-115 (Fig. 6, bottom left) and 120 go together, as Bentz saw, in terms of fabric, paint and shape, as do 113 and 114 (Fig. 6, top), which have the same polished surface as C-1937-2469. Generally, in these vases, the body is increasingly ovoid, the shoulder less rounded, the lower wall straighter; the neck becomes taller in relation to the body; the ridge or drip-ring between neck and shoulder is now omitted; and the handle is increasingly concave on its upper surface. Most significant is the base, for the ring of the foot is broad and barely offset from the underside (false ring-foot), that of 115 and 120 representing almost a return to the type of foot used in Early Corinthian for such vases as C-1962-569.¹⁹ The two bands on the body are now thinner and closer together than on oinochoai of the sixth century. The increasing height of the neck in relation to the body is emphasized in 116 (Fig. 6, bottom right), which Bentz is probably right in considering the latest Corinth oinochoe from this well, and the tall ovoid body has a flatter shoulder than before. A new feature on 116 is the addition of a band on the inside of the neck, just below the rim.

Figure 6.

One-handled oinochoai from well 1939-1. Scale 1:3

Figure 7.

One-handled oinochoai from pit 1975-1. Scale 1:3

C-1975-135 (Fig. 7, top) and C-1975-308 (Fig. 7, bottom), from the so-called Amphora Pit deposit (pit 1975-1) and datable about 480–440, may have been made in one workshop, given their similar fabric and potting. But, assuming that they are contemporary expressions of the one-handled oinochoe, they nevertheless reveal some differences in shape. Both have an ovoid body with high maximum diameter, and both omit the ridge at the base of the neck, but 135 has a taller neck than 308, and a flat base in contrast to the strong ring-foot of the latter. The rim of 135 projects and is rounded on top, while that of 308 is thickened and has a shallow groove. Both vases are decorated similarly with bands, except that 135 has an additional band on the inside of the neck and another on the beveled edge of the base. In shape (including the form of handle), as well as in the potting, C-1975-135 is very similar to C-1939-113 and 114 (Fig. 6, top), which may be approximately contemporary.

The third quarter of the fifth century seems to have been a crucial time in the development of the Corinth oinochoe, at least to judge from the evidence provided by such deposits as well 1934-10 ("Pease Well"), which was filled, perhaps on a single occasion, with pottery mainly of the years 460–420. Among the fragmentary vases were three certain one-handled oinochoai: C-1934-949 (Fig. 8, top left), C-1934-950, and C-1934-1193 (Fig. 8, bottom). Of these three, only the first preserves the complete profile (minus handle and most of the base), but enough remains of the other two to show that all three were similar to C-1975-135 in body and neck, and the partially preserved handles of C-1934-950 and 1193 retain the type used for C-1975-135. Whether the feet were treated in a similar manner is uncertain.²⁰ The rim is now broader (ca. 1 cm) than that of C-1975-135, which would be appropriate if these three vases are indeed a little later in date, from the third quarter of the fifth century. In her publication of this well, Pease classes the more fragmentary oinochoe C-1934-951 (Fig. 8, top right) with 949, but a close examination of the profile shows that significant differences exist: 951 has a much lower neck and a slightly outturned rim, rounded rather than flat on top. Nor is it certain whether 951 had one or two handles. Whatever the case, 951 is a bridge between the one-handled form represented by C-1934-949 and contemporary two-handled oinochoai such as C-1975-307 (see below, p. 55 and Fig. 14, top).²¹

Figure 8.

Oinochoai from well 1934-10. Scale 1:3

As we will see (below, pp. 56–57), well 1934-10 contained at least as many, if not more, two-handled as one-handled Corinth oinochoai. This is symptomatic of a decline in the production of the one-handled Corinth oinochoe in favor of the two-handled version. Although the presence of one-handled oinochoai in well 1934-10 indicates that the version was still in production during the third quarter of the fifth century, it is not clear when the form disappeared from the repertory of the Corinthian potter. In this context, we must consider the jug C-1937-410 (Fig. 9, left), which comes from well 1937-1 and so should be dated to the last quarter of the fifth or the first quarter of the fourth century. This vessel certainly had only a single handle and, although the attachment at the rim is not preserved, the handle would have been difficult to grasp unless it rose high. But the globular shape and low, outturned rim are unusual if we compare examples from pit 1975-1 and well 1934-10. Presuming that the pot is Corinthian, it represents something of an anomaly at the end of the series.²² Although no other complete examples of the one-handled oinochoe that are certainly later than ca. 420 have been inventoried at Corinth, there are, however, some sherds from contexts of the last quarter of the fifth and the early fourth century. For example, drain 1937-1, the pottery of which belongs mainly to the first quarter of the fourth century, included the rim and high-swung handle of a one-handled oinochoe (L-1937-3-1). A rim and high handle (L-1972-55B-1; Fig. 9, right) from a similar vase and the rim of a possible second example come from lot 1972-55B, in which the bulk of the pottery belongs to the fourth century, but isolated pieces from this context are as early as the first half of the fifth.²³ To be sure, these and other instances may all be dismissed as heirlooms or as earlier pieces that have strayed into deposits that contain mainly later material, but we cannot completely rule out the possibility that the occasional one-handled oinochoe was still potted in the late fifth or even in the early fourth century.

Figure 9.

One-handled oinochoai from various contexts. Scale 1:3

Figure 10.

Two-handled oinochoe. New York, Metropolitan Museum of Art 1976.223. Courtesy Metropolitan Museum of Art, New York. Scale 1:3

The Two-Handled Corinth Oinochoe ("Decanter")

Archaic

The earliest extant Corinth oinochoe with two handles is today in the Metropolitan Museum of Art in New York, inv. no. 1976.223 (Fig. 10).²⁴ The ring-foot is broad, with a sloping edge; the bulbous body has a relatively low point of greatest diameter; the neck is low, flaring, and ends in a flat rim; and two flattened oval handles are set opposite each other. The edges of the foot and the mouth are black, and a broad band on the body acts as the ground for a scene of two confronting padded dancers on each side. In terms of shape, and indeed in overall dimensions, the vase is very close to the earliest form of the one-handled Corinth oinochoe, especially C-1932-134 (Fig. 2), which is also decorated with padded dancers.²⁵ As we have noted, Amyx and Lawrence consider C-1932-134 to be Early Corinthian, and on grounds of style, Amyx has suggested a date for the New York oinochoe of "ripe EC/early MC," so perhaps ca. 610–585 on his chronology.²⁶ Although the New York oinochoe may not be quite as early as the earliest preserved Corinth oinochoai with one handle, nevertheless it clearly shows that the two-handled variety was introduced not much later than the one-handled form, if not at the same time—on present evidence possibly within the EC period, in the last quarter of the seventh century.²⁷

Figure 11.

Two-handled oinochoe C-1937-2056 from well 1937-3. Scale 1:3

Despite the emergence of the two-handled oinochoe in Corinthian pottery during the late seventh or early sixth century, examples from the sixth and first half of the fifth century are uncommon, and it is not possible to trace a general development in shape.²⁸ The earliest example from the American excavations in Corinth is C-1937-2056 (Fig. 11), which was found in well 1937-3 and should, therefore, belong to the second half of the sixth century.²⁹ This is the first of Edwards's "short-lived experiments."³⁰ The foot here takes the form of the bases of most of the one-handled Corinth oinochoai from the same context, a simple disk, very slightly concave underneath. The body is globular and not unlike C-1939-119 in profile. The slightly flaring neck ends in a rounded lip.³¹ The high-swung handles are not set opposite one another as on the earlier New York oinochoe but about 100° apart, and are approximately oval in shape. The decoration has been abraded, but seems to have consisted of one or two horizontal bands above and below the greatest diameter, and, unusually, a short vertical wavy or zigzag line on the shoulder, perhaps a last remnant of the shoulder decoration of earlier Corinth oinochoai.³²

Figure 12.

Two-handled oinochoe C-1967-129 from well 1936-10(?). Scale 1:3

Classical

Before we examine the Classical series of Corinthian decanters, we need to look at a two-handled oinochoe that is something of an anomaly, namely, C-1967-129 (Fig. 12). It shares with C-1937-2056 from the Campbell Well a slightly concave base, and flattened oval handles (2.4 cm wide) set about 100° apart. But there are basic differences: the body is biconical; the neck broad and low; the rim is everted, flat on top and rounded at the edge; and, in particular, the low handles do not rise above the rim. A little above the greatest diameter, a broad black band is flanked by two narrower purplish-red lines. The vase has no definite context, but one may conjecture that it perhaps came from well 1936-10. In her article of 1937 publishing the pottery from well 1934-10, Pease discussed a number of decanters from contemporary deposits, noting that "from the aforementioned well at S:11 [well 1936-10], South Basilica, came two vases with two handles on the same side of the vase. They are squatter; one is unglazed, the others [sic] glazed in the 'conventionalizing' style."³³ The squat shape and "conventionalizing" decoration of red and black lines suggest that C-1967-129 may be the missing decanter from this well. If this is correct, the pottery from well 1936-10 would suggest a date in the first three quarters of the fifth century.³⁴ Otherwise, the only means of dating is the combination of black and red bands. This is a form of "conventionalizing" decoration which is used on a variety of shapes from the middle of the sixth to the second half of the fourth century.³⁵ A date in the second or third quarter of the fifth century for C-1967-129 would be appropriate for both the putative context and the decorative motifs.³⁶ In any case, this unusual jug may be an experiment, utilizing the low biconical body of a trefoil broad-bottomed oinochoe, or may even be an imitation of a non-Corinthian form.³⁷

With C-1939-22 (Fig. 13), from the first half of the fifth century, we begin a continuous series of two-handled Corinth oinochoai.³⁸ C-1939-22 was found in well 1939-1 with the eight one-handled oinochoai C-1939-113 to 120 discussed above (p. 47). It has a flat base similar to that of C-1937-2056 but taller and with a more beveled edge. The body is more ovoid, with high maximum diameter, and the rim is more outturned, while retaining its simple rounded form. The handles still curve high above the rim, but are now straplike with slightly flanged edges and are set somewhat closer (about 90° apart). The decoration consists of bands on the edge of the foot, above and below the maximum diameter, at the rim, and on the inside of the neck: this combination is obviously borrowed from the contemporary one-handled oinochoe, and will become standard. Although there are some differences in detail (lower neck, form of rim and foot), C-1939-22 resembles, both in general shape and in potting, the one-handled oinochoai C-1939-113 and 114 from the same deposit, to such an extent that all three probably come from the same workshop. In her analysis of the Corinth oinochoai from this deposit, Bentz places these three vases in the second quarter of the fifth century.³⁹ This may be correct, and certainly a more general date in the first half of the fifth century seems warranted.⁴⁰

Figure 13.

Two-handled oinochoe C-1939-22 from well 1939-1. Scale 1:3

The next stage in the development of the two-handled oinochoe is represented by C-1975-307 (Fig. 14, top) and L-1975-132-23 (Fig. 14, bottom), which were found with the one-handled oinochoai C-1975-135 and 308 in pit 1975-1, and may be dated about 480–440. C-1975-307 and L-1975-132-23 must again be the products of a single potter. C-1975-307 has flanged handles set about 80° apart, a little closer than on C-1939-22, but still looping above the mouth. The false ring-foot is similar to the feet of the one-handled oinochoai C-1939-113, 114, and 116, of the second quarter and middle of the fifth century. The real change, in respect to the slightly earlier C-1939-22, has come about in the upper half of the vase: there is a continuous curve from the simple outturned rim through an almost nonexistent neck to a concave shoulder that passes into the convex lower body. The result is a squatter vase with smaller mouth in relation to the greatest diameter.⁴¹ L-1975-132-23, of which only one handle remains, was probably a two-handled oinochoe. Apart from the rim, which is not preserved, the shape is the same as that of C-1975-307, even to the form of the base. The decoration of bands is also the same on the two vases, so far as one can tell. The form of body that we see in these two vases was not entirely new, but was a development from the shape represented by the one-handled oinochoai C-1937-1017 (Fig. 4, center left) and C-1939-117 (Fig. 5, left), from the last half of the sixth century.

Figure 14.

Two-handled oinochoai from pit 1975-1. Scale 1:3

The most important evidence for the two-handled oinochoai during the third quarter of the fifth century comes from well 1934-10. There are at least six fragmentary examples: C-1934-946 (Fig. 15, top left), 947 (Fig. 15, top right), 948, 1194, 1195 (Fig. 15, bottom), and 1196.⁴² The first four have a body that derives from the type represented by C-1975-307, although the neck is more pronounced, but the point of maximum diameter is lower and clearly articulated (Edwards's "shoulder stop") at the junction of shoulder and body, at least in the case of 946 and 947; in 948 and 1194 the transition is more gradual.⁴³ The bases of 946, 947, 1195, and 1196 are similar to that of C-1975-307, except that in the case of 1196, the broad groove offsetting the false ring-foot from the underside has been reduced almost to an incised line. The lip is still outturned but is now flat or slightly concave on top, not unlike the lip that we find on certain one-handled oinochoai (C-1934-950, 1193) from the same deposit. The main change, however, has come in the form of the handles: they are still set apart approximately 80° (C-1934-946, 1194) to 100° (C-1934-947), but instead of being concave on the outer surface, with flanged edges, they are now convex, with central rib;⁴⁴ furthermore, the handles now rise only to the level of the rim, or a very little above. This low handle, with slight variations, will remain the preferred type from now on. The syntax of bands remains unchanged, but some variation occurs in the area of the mouth: the band on the rim may pass across the handles (946, 1194), or stop (947, 948), leaving a reserved space, and the band on the inside of the neck may be included (946, 948) or omitted (947, 1194). These changes produce Decanter type I, in Edwards's typology.⁴⁵ Now, Edwards considered that the earliest examples of his Decanter type II also came from this deposit, namely, C-1934-1195 (Fig. 15, bottom) and 1196.⁴⁶ In both cases, however, only the base and lower body remain. The type of base and the profile of 1195 suggest that it comes from an oinochoe like C-1934-947 (i.e., Decanter I). On the other hand, as Edwards maintained, 1196 certainly has a flatter shoulder and squatter body, but as the squatter body came to be used, as we will see, with rims of various types, it is really not possible to be certain how the missing parts should be reconstructed.

Figure 15.

Two-handled oinochoai from well 1934-10. Scale 1:3

In summary, we may say that the Classical series of two-handled oino-choai begins in the first half of the fifth century, perhaps in the second quarter. At first there is some diversity of shape, as illustrated by C-1939-22 and C-1975-307, but the interconnection of the two-handled and the one-handled forms is quite clear, thus confirming the relationship suggested by Lawrence, and later by Bentz.⁴⁷ By the third quarter of the fifth century, as evidenced by well 1934-10, production of the two-handled oinochoe was increasing, and the one-handled variety was losing favor. It is not possible to know what may have stimulated this change, but one may perhaps conjecture that it was connected with increasing wealth in Corinth after the Persian Wars and with consequent changes in sympotic fashion in the city.⁴⁸ There is also ample evidence demonstrating that, during the second and third quarters of the fifth century, there was considerable experimentation by the Corinthian potters with the introduction of remodeled or new shapes as well as new modes of decoration. It was at this time that the old forms of Corinthian transport amphorae, types A and B, were re-worked, and a new type, AΕ, introduced.⁴⁹ This was also the time of the Outline Style, the Sam Wide vases, and the earliest Corinthian versions of white-ground, red-figure, and stamped black-glaze pottery.⁵⁰

Figure 16.

Two-handled oinochoe L-1981-71-8 from well 1981-2. Scale 1:3

In one important context of the third quarter of the fifth century, the lower filling in well 1981-2, the forms represented by C-1975-307 (Fig. 14, top), on the one hand, and C-1934-946 and 947 (Fig. 15, top), on the other, are found together. L-1981-71-9 preserves part of the high flanged handle and rounded outturned lip (Diam. 9–10 cm) of an oinochoe like C-1975-307. On the other hand, L-1981-71-8 (Fig. 16) comes from an oinochoe like C-1934-946 and 947, with low handle ridged on the upper side; but at the junction of shoulder and neck, we have the unusual addition of a groove between two ridges.⁵¹ In addition, the lot has also produced a downturned rim (L-1981-71-12) which may come from an example of Edwards's Decanter II. It is clear, therefore, that during this period of transition for the Corinth oinochoe, the potters were experimenting with many elements of the shape. Thus, the earlier flat rim may be downturned, or downturned and thickened, and it is not surprising that forms of rim that are transitional between Edwards's Decanters I and II (see Fig. 17) are found in other deposits of the third and fourth quarters of the fifth century.⁵²

Figure 17.

Two-handled oinochoai from various contexts. Scale 1:3

At the same time that these new types of rim evolve, the body of the decanter begins to become even squatter. From lot 7193, connected with Building II and dating ca. 425–375, L-7193-4 (Fig. 18, top) preserves a thickened rim and retains a good part of the body.⁵³ The profile may be compared with that of C-1934-946 (Fig. 15, top left), but the shoulder is a little flatter and the body must have been squatter. In fact, the general shape was similar to the more complete jug C-1931-66 (Fig. 18, bottom left), classed as a Decanter II by Edwards and dated ca. 425, although it cannot be placed more closely than ca. 425–325 from its context.⁵⁴ Here the rim is downturned, although not thickened. The handles have the ridged form of C-1934-946 but are set closer together (now about 50° apart) and join the base of the rim, so not looping as high. The false ring-foot resembles the earlier C-1975-307 and C-1934-946, but the body is noticeably squatter and less tall. Very similar to C-1931-66 in the form of the rim and handles are L-7193-5 and L-2568-1 (Fig. 18, bottom right), the latter from the Vrysoula deposit (channel 1964-1) and datable to the last quarter of the fifth century.⁵⁵

From this discussion, it will be apparent that the development of the two-handled Corinth oinochoe in the last half of the fifth century was more complex than previously thought: different types of rim were in use at one time; there was a shift to a ribbed handle, gradually attached lower on the rim; and the body became increasingly squatter. It should also be noted that some of these elements—the thickened rim used for type II decanters and the ribbed handle found on both types I and II—were also used for imitation blisterware amphorae of the last half of the fifth and first half of the fourth century, which may well have been made in the same establishments.⁵⁶

Figure 18.

Two-handled oinochoai from various contexts. Scale 1:3

By the last quarter of the fifth century, the two-handled jug must have become one of the major forms, if not the major form, of fineware pouring vessel in Corinth.⁵⁷ For the late fifth and first quarter of the fourth century, perhaps the two most important deposits are well 1937-1 and drain 1937-1. The pottery from well 1937-1 belongs to the last quarter of the fifth century and to the first quarter of the fourth, and includes two fragmentary two-handled oinochoai: C-1937-451 (Fig. 19) and C-1937-454. C-1937-451 preserves a complete profile, but many parts are missing, and it is not clear whether the base was flat or not. Of 454, only the upper half remains, which does at least show that the vase was substantially larger than 451. Both 451 and 454 have the thickened, downturned rim characteristic of Edwards's Decanter II, slightly convex on the outer face, and undercut in the case of 454. The handles are strongly ridged, and rise just above or just below the level of the mouth, but they are now set closer together (about 50° apart, like those of C-1931-66) than on the jugs from well 1934-10. Quite distinctive is the indentation on the inside of the neck just below the rim. There is now a continuous curve from upper neck to foot, and the body of 451 is low and squat. Both 451 and 454 have bands on the outer edge of the rim that cross the handle-attachments, and on the inside of the neck, and 451 has two bands on the body, one at the point of greatest diameter, the other just below the lower handle-attachments, the two bands closer together than in the oinochoai from well 1934-10.

Figure 19.

Two-handled oinochoe C-1937-451 from well 1937-1. Scale 1:3

The pottery from drain 1937-1 is largely contemporary with that of well 1937-1 but may continue into the early second quarter of the fourth century. From the drain, nine decanters have been inventoried (C-1937-240 to 248); C-1937-242 (Fig. 20, top) and C-1937-243 (Fig. 20, bottom) illustrate the examples.⁵⁸ All display the same distinctive characteristics in shape, fabric, and potting, so that they must come from a single workshop, as Kazazis, Morris, and McNiven pointed out in an unpublished study of the pottery from the well and drain.⁵⁹ The squat body is similar to that of C-1937-451 from well 1937-1; the false ring-foot is like that used earlier for C-1975-307 and C-1934-946, but the groove separating the narrow outer ring from the slightly convex underside of the base is broader and deeper. The edge of the overhanging rim may be straight or slightly concave, not convex as in C-1937-451, although the same shallow indentation on the inside of the neck is present in all. The handles attach to the bottom of the rim, like those of C-1937-451 and 454, but they do not curve up as much, are set slightly closer together (ca. 45° apart), and, most significantly, they do not have a central rib but are flanged, rather like the earlier type represented by C-1975-307. The syntax of bands, different from that on the examples from well 1937-1, is standard on all nine decanters from the drain: a broad band on the edge of the foot; a broad band below the point of greatest diameter; two narrower bands on the shoulder; a band (not broken by the handles) on the edge of the rim; and two bands on the inside of the neck.

Edwards had no evidence for his Decanter type I between the third quarter of the fifth century and the Hellenistic period. It is now clear, however, that types I and II were both in use during the last quarter of the fifth century and the first quarter of the fourth, along with rim forms

Figure 20.

Two-handled oinochoai from drain 1937-1. Scale 1:3

that may be regarded as transitional between the two types. From drain 1937-1, in addition to the nine type II decanters, there are two rims (L-1937-2-35, L-1937-3-3) that are slightly convex on top and outturned, illustrating the transition from type I to II. Other deposits reveal the same situation. C-1972-92 (Fig. 21, top) must be approximately contemporary with the oinochoai from well 1937-1. It has the same low, squat body and tall, narrow neck as C-1937-451, a type II decanter, and the organization of bands is similar. The base is flat, the underside slightly concave. Most importantly, however, the everted rim of C-1972-92 is a thickened version of the type I decanter as represented by C-1934-947 (Fig. 15, top right).⁶⁰ Moreover, the ribbed handles, set about 50° apart, rise somewhat above the level of the mouth and do not attach at the bottom of the rim as do those of the type II decanter.⁶¹ The same combination of type I and II rims is found in pit 1972-1, where the pottery covers the fourth century down to ca. 320 or a little later. C-1972-117 (Fig. 21, bottom left) and L-1972-63-6 have rims of type I, developed from the C-1934-946 type. At least three other oinochoai from the deposit (L-1972-63-5 [Fig. 21, bottom right], L-1972-63-7, and one unlotted) had the more downturned and thickened rims of type II, and to judge from the potting and banding of L-1972-63-5 and 6, the same potter was producing decanters with type I and II rims. The handles of all four pots are ridged and set about 50° apart. Three additional, uninventoried bases in the deposit, together with C-1972-116, are flat, the underside slightly concave.⁶² It thus seems that types I and II, with their variations, represent alternate types of rim for flat-based, squat-bodied decanters from the third quarter of the fifth century into the fourth century, possibly through the second quarter.⁶³

Figure 21.

Two-handled oinochoai from various contexts. Scale 1:3

Oinochoai of all three of Edwards's types are found in lot 1978-53, an important fill, connected with the destruction of Building III, that seems to have been dumped during the last quarter of the fourth century. L-1978-53-6 (Fig. 22, top left), although missing much of the body and neck, is clearly an example of type I. It has a false ring-foot marked off by a shallow groove; a body less squat than in type II; handles oval in section,

Figure 22.

Two-handled oinochoai from lot 1978-53. Scale 1:3

attached to the rim about 90–100° apart and rising a little above it; and a rim that flares out, is almost flat on top, and has a rounded edge. L-1978-53-8, which preserves parts of the neck and lip, is so similar that it must come from a second type I decanter that was produced in the same workshop. Both jugs have a broad band on the top of the lip and a second band, well down on the inside of the neck.⁶⁴ In contrast, C-1978-39 (Fig. 22, top right), although missing the base, clearly belongs to type II and may be compared with C-1937-451 (Fig. 19) and 454 from well 1937-1: the ridged handles, set about 50° apart, curve over to join the lower edge of the lip in a similar way; the lip is modeled and undercut; the neck has a similar indentation below the rim on the inside; the body is low and squat.⁶⁵ Now, Edwards thought that the latest preserved examples of his Decanter type II had to be dated in the early first quarter of the fourth century, although he conjectured that the type may have continued into the second quarter. He also believed that Decanter type III developed out of type II during the second quarter of the fourth century, even if the earliest preserved examples were to be dated to the early third quarter. But the evidence from these two deposits, as well as that from drain 1971-1 to be considered below, suggests that decanters with type II rims continued to be produced probably into the third quarter of the fourth century. Significantly, lot 1978-53 also contains one example, L-1978-53-7 (Fig. 22, bottom), of Edwards's type III, including parts of the mouth, neck, and handles. The tall, straight neck, vertical collar rim, and ridged handles set about 40° apart and reaching up to the bottom of the collar are all characteristic of type III. The date of L-1978-53-7 is not certain, but it is unlikely to be earlier than the third quarter of the fourth century. Such a date is supported by the contexts of some of the six type III decanters listed by Edwards: C-1937-2521 (Fig. 23) and 2662 come from pit 1937-1, which provides a date in the third quarter or early fourth quarter of the fourth century; C-1953-60, of which the neck, mouth, and handles are missing, comes from "Grave" 1953-6, which has a terminal date of ca. 320, similar to that of pit 1937-1.⁶⁶

Figure 23.

Two-handled oinochoe C-1937-2521 from pit 1937-1. Scale 1:3

What, briefly, are the characteristics of the type III decanter? If we look more closely at C-1937-2521 (Fig. 23) and 2662, C-1953-60, and C-1931-238, we see that, where the base is preserved, it is flat or slightly concave underneath. The body now is less squat, with a higher maximum diameter, so that the lower wall is more diagonal. The neck is taller and narrower, as is the mouth, to which a thin strip of clay is added to form a "collar rim," to use Edwards's evocative term. The handles are now taller, join at the bottom of the rim, and their central rib is less emphasized. They are usually placed about 50–60° apart (50° for C-1937-2521 and 2662, 60° for C-1931-238). Two bands continue to be applied in glaze on the body just above or below the maximum diameter and on the shoulder, as well as on the edge of the collar rim and on the inside of the neck, but not on the edge of the base.

Of all the pre–South Stoa deposits, drain 1971-1 is the richest in two-handled oinochoai (see Fig. 24). On the basis of handles and distinctive shoulders, this one deposit contained at least 18 decanters. Since the pottery from this drain includes shapes suitable for the storage, preparation, and serving of wine, along with the preparation and eating of food, and must come from one or more establishments with dining facilities, the quantity of decanters clearly demonstrates the significance of the shape in dining in the second half of the fourth century. Disk-feet with concave underside and rounded or tooled edge are used for all identifiable bases. The body, where preserved, is that normal for a decanter of type III, but the neck may be straight and cylindrical (C-1971-236 [Fig. 24, bottom left], L-7079-210) or may flare out toward the top (L-7079-206 [Fig. 24, top], 211). There may be a tendency for the flaring neck to be associated with the undercut collar rim (L-7079-206, 211) and for the straight neck to go with the fused collar rim (C-1971-236, L-7079-210). Of the two forms of rim, the strongly undercut is the more common, and one wonders whether the fused form, in which the rim is joined to the upper neck, is not a slightly later development. Although the collar rim, whether undercut or fused, is thus the predominant form, there is at least one belated example (L-7079-208) of a thickened type II rim. Of most interest, however, is L-7079-209 (Fig. 24, bottom right), which preserves a complete profile except for the lower wall and base. The flat shoulder and diagonal lower wall give a more pinched shape to the body, the ridged handles again rise to the level of the mouth as in earlier decanters of type I, and the outturned rim is not fused or undercut like the normal type III but looks like an unusually thickened form of type II. The two bands on the body are broader than usual and close together at the point of greatest diameter, and the band on the inside of the neck is also unusually wide. It is possible that L-7079-209 may represent the latest form of decanter in the drain, perhaps datable to near the end of the fourth century.⁶⁷

Figure 24.

Two-handled oinochoai from drain 1971-1. Scale 1:3

Hellenistic

The construction of the huge South Stoa, and the introduction of new shapes, especially forms of kantharoi, along with "West Slope" decoration, mark the beginning of the Hellenistic period archaeologically at Corinth. Of the decanters classed as Hellenistic by Edwards,⁶⁸ C-1931-280 (Fig. 25, left) is closest to the form favored in the later fourth century. However, it has a more diagonal shoulder and generally a rather biconical body; and three, not two, bands on the body, rendered in a thin brown slip. The handles (55–60° apart) have also lost the central rib. The clay is rather coarser, has fired harder, and is a greenish yellow (2.5Y 8/4–6), unlike the fabric of jugs in the fourth century. The vase comes from well 1931-8, but the context is unhelpful, and a date anywhere from the late fourth century (i.e., before the construction of the South Stoa) to the later third is theoretically possible.⁶⁹ Certainly the vase has more in common with late fourth-century decanters than with those of the Hellenistic period from Corinth.

The second of Edwards's Hellenistic decanters is C-1940-413 (Fig. 25, right), an unusually tall example of the shape. Edwards dated this vase to the first quarter of the third century, like C-1931-280. This dating was based partly on shape development, partly on the context, well 1940-1, which included pottery from the last quarter of the fifth century to ca. 275. Although the final date for the closing of the well should perhaps be lowered to about 250, as Pemberton has suggested for the contemporary filling of cistern 1940-1,⁷⁰ a date for C-1940-413 at the end of the fourth or in the first half of the third century is likely. C-1940-413 resembles C-1931-280 in the disk-base; biconical body (though here more pronounced); tall, narrow neck; and flat oval handles, though these are set further apart (ca. 75° rather than 50°) and, most unusually, are attached on the neck below the lip. The decoration of five bands on the body also diverges from what is traditional for a Corinthian decanter. The fabric of C-1940-413 is a pale brown, close to 7.5YR 7/6, fine but with much fine silver mica. This, combined with the other unusual elements in the shape and decoration, suggests that C-1940-413 may not be Corinthian. Thus, the evidence for the continuation of the traditional form of the two-handled oinochoe at Corinth into the third century is not great, and it is possible that the form may have disappeared by about 300.

Although the old form goes out of fashion, two-handled jugs continued to be made at Corinth at least during the third century and possibly down to 146, but in quite different forms and certainly in significantly reduced numbers. Some variation in shape is noticeable, but essentially we see the reintroduction of a plainer, more utilitarian type; and the shape is now made in cooking ware, at times with an indented base. These two-handled jugs of the Hellenistic period do not seem to have been primarily vessels for pouring wine at the symposion, but there is no Corinthian shape that clearly takes up this function. Edwards noted the absence at Corinth of decorated jugs in the Hellenistic period and pointed out that the fineware jugs were few in number and mainly of small size.⁷¹ He suggested that "in simpler households the need for larger wine pitchers may have been supplied by examples in coarse ware . . . or by such shapes in cooking ware," and this may be correct, but he was understandably hard put to explain this phenomenon. This situation seems to mirror the virtual disappearance of the krater from the ceramic repertory at Corinth and in other mainland Greek centers during the Hellenistic period.⁷² It may be conjectured that, after Alexander's conquests, which seem to have increased the availability of metalwork in Greece, metal vases took the place of the ceramic krater and oinochoe at the symposion, but, as Susan Rotroff has argued, the evidence perhaps points to a more radical change in drinking customs.⁷³ The symposion and the sympotic krater were closely connected with male society in the independent poleis of the Classical period. The Macedonian hegemony, which brought an end to the independence of city-states such as Corinth, may well have initiated, or at least hastened, profound social, as well as political and economic, change.

Figure 25.

Two-handled oinochoai from various contexts. Scale 1:3

The change from the decanter of the Classical period is especially marked in the remaining Hellenistic vessels to be discussed here. C-1975-283 (Fig. 26) is made from a fine, pale brown clay and is quite thin-walled (ca. 3–5 mm). It has a stretched biconical body that passes into a tall, narrow neck, and a broad projecting rim with rounded edge. The handles, which are flattened ovals and set about 50° apart, join the top of the neck, pushing up the rim a little. The jug has a very narrow ring-foot with nippled underside. This form of base is replicated on a contemporary Corinthian lagynos, C-1981-113.⁷⁴ The modest decoration is quite different from that of the earlier series of two-handled jugs: on the neck are two horizontal grooves about 8 mm apart; and the upper part of the neck, the lip, and the tops of the handles have been dipped in a brown to light red slip. C-1975-283 came from the fill of well 1975-5. The pottery in the well goes down to the middle or the third quarter of the third century, so the jug probably belongs to the first half or, at most, the first three quarters of the third century.

Figure 26.

Two-handled oinochoe C-1975-283 from well 1975-5. Scale 1:3

C-1947-314 (Fig. 27, top left) and C-1981-45 (Fig. 27, top right) have a similar profile, although the former is taller and has a narrower neck. Each has a narrow, essentially flat base, marked off from the body of the pot only by a slight inward curve of the wall, but there is no separate foot. An ovoid body rises from this base with continuous profile through the short neck to the rim, a form quite different from the bodies of two-handled jugs in the fourth century. The lip has none of the modeled forms of the fourth century, but flares out to a simple rounded rim. Two handles, set about 55–60° apart, are now flattened ovals without any central ridge, and rise to the level of the rim or very slightly above it, attaching to its edge. In these two vases, no painted decoration is present. They are both wheelmade, with smoothed surfaces, in a fine, moderately hard fabric.⁷⁵ C-1947-314 was, of course, known to Edwards, who postulated that it might represent "one of the latest stages in a direct shape development" from his type I decanter of the third quarter of the fifth century,⁷⁶ although he fully recognized that this remained to be demonstrated, since no intermediate examples were known. It is, however, preferable to regard the form simply as a return to a plainer jug after the more extravagant varieties of the fourth century.

Now, C-1947-314 came from the fill of well 1936-12. Unfortunately, the pottery covers the whole period from the late fourth century to 146. Edwards, however, dated this vase to the first half of the second century, on the basis of its resemblance to the two-handled jug C-1947-829 (Fig. 27, bottom left), from well 1947-3 in the Southeast Building, a context that he felt belonged to the time of the destruction of Corinth. This deposit has recently been published fully by Irene Romano, who argues that the principal filling was probably dumped in the late second or early first century B.C.⁷⁷ While most of the pottery belongs to the first half of the second century, much of it is earlier. Therefore, we cannot rule out an earlier date, in the third century, for C-1947-829 and also C-1947-314. In fact, such a date is strongly suggested by the context of C-1981-45. This also came from a well, 1981-2, in the excavations east of the Theater. The upper fill (lot 1981-70), which included C-1981-45, clearly belongs to the first half of the third century. Therefore, it may be that all three of these two-handled jugs are to be dated to the first half of the third century. Certainly, the broad similarity of potting between C-1947-314 and C-1981-45 may indicate that they were produced in the same workshop.

Figure 27.

Two-handled oinochoai from various contexts. Scale 1:3

C-1947-829 (Fig. 27, bottom left), which we have just considered for its context, has essentially the same shape as C-1947-314 and C-1981-45, with handles set also about 55–60° apart, but which join just below the rim, the poorly set right handle rising somewhat above it. The vase is wheel-made, with horizontal shaving on the lower body. In this case, however, we have, for the first time, a two-handled jug manufactured in a coarse fabric.⁷⁸ C-1987-78 (Fig. 27, bottom right) has a similar shape: globular body with continuous curve through a short, broad neck to a simple everted lip; flat oval handles attached to the rim about 60° apart. There is no foot, but the base has a broad resting surface with concave underside and central nipple.⁷⁹ The vase belongs to the fill of cistern 1987-1, and probably dates to the third century or to the first quarter of the second. A very similar two-handled oinochoe, IP 563, comes from the upper fill of the South Slope cistern in the Rachi settlement at Isthmia, where the context indicates a date in the second half of the third century.⁸⁰

Figure 28.

Two-handled oinochoai from manhole 1981-6. Scale 1:3

Another deposit east of the Theater, manhole 1981-6, has produced at least three two-handled jugs in typical cooking ware: C-1981-115 (Fig. 28, top), L-1981-108-2 (Fig. 28, bottom), and L-1981-108-3. Unfortunately, the context provides no closer dating than ca. 300–146. All three vessels are wheelmade and thin-walled.⁸¹ Both of the illustrated vases have a simple flaring lip and narrow strap or flattened oval handles, set about 50° apart, that rise from the shoulder to join the rim. The body is globular but is clearly offset by a jog from the tall, conical neck. The most unexpected feature is the strong indentation of the bottom of each of the three oinochoai. This feature indicates that these vases are really two-handled versions of the Hellenistic round-mouthed pitcher.⁸² Edwards noted that the round-mouthed pitcher was employed over the fire for boiling, in addition to its primary function in collecting and carrying water. There is no certain evidence of secondary burning on the three two-handled jugs from manhole 1981-6, but it is difficult to see why they would be made in cooking ware unless they were intended to be used over a fire or at least to hold heated contents.

Figure 29.

Two-handled oinochoe C-1981-102 from manhole 1981-4. Scale 1:3

The latest of our two-handled jugs from Corinth may be C-1981-102 (Fig. 29). It has an offset, flat base, but quite low, globular body that passes into a narrow neck, and a flaring rim that is grooved on the outer edge as well as on top. The narrow strap handles, each with two ridges, are attached to the top of the neck at the lower part of the rim. The vase is wheelmade; the fabric moderately hard, coarse, and with small to large grits, both dark and white, the core and surfaces grayish brown (10YR 6–7/4). The fabric is possibly, but not certainly, Corinthian. The context, the fill of manhole 1981-4, had some earlier material, but most of the pottery belonged to the later first century B.C. Since the jug came from the very top of the fill, with earlier material, one cannot be sure whether it should be assigned to the Roman period (i.e., after 44 B.C.), or is earlier, perhaps from the interim period between 146 and 44.

Corinth and Athens

A high-handled, round-mouthed jug, with black neck and band on the body, began to be manufactured in Athens from about 600 B.C.⁸³ Sparkes and Talcott assert that it was adopted from Corinth, being an adaptation of the one-handled Corinth oinochoe, but as Lawrence has rightly remarked, the earliest Athenian form is quite different, particularly in details of body and neck, from the Corinthian, so that the connection is by no means assured.⁸⁴ It is, however, true that there is a closer relationship between the two shapes from the second half of the sixth century, as Lawrence points out, and this is particularly so toward the end of the Athenian series in the first half of the fifth century. This can be seen if Agora P 16506 is compared with C-1939-116 (Fig. 6, bottom right) or C-1975-135 (Fig. 7, top).⁸⁵ One-handled Corinth oinochoai, then, may have been occasionally reaching Athens, although I am unaware of any specific imports.

It is also in the first half of the fifth century, about 480–470, that two-handled jugs appear at Athens in the hands of symposiasts, on two cups painted by Douris or a follower.⁸⁶ An exceptional shape clearly caught the eye of the painter, and was recorded. Now, there is no evidence for the manufacture of ceramic two-handled jugs in Athens at so early a date, so that the originals must have been of metal or, if they were clay, they must have been imported, whether from Corinth or another center.

During the fifth and the fourth century, particularly the latter, the two-handled jug was a common shape in many Greek centers throughout the Mediterranean, often in quite individual forms.⁸⁷ It is unnecessary to explore these variations in this place, but I would like to consider a little further the particular connection between Corinth and Athens. During the Classical period, Corinthian pottery, both fine ware and coarser domestic ware, reached Athens in considerable quantity.⁸⁸ Typical Corinthian two-handled jugs were exported to Athens and elsewhere in Attica during the last quarter of the fifth and the first quarter of the fourth century.⁸⁹ But, toward the beginning of the fifth century, potters in Athens had introduced their own peculiar type of round-mouthed jug, characterized by a convex upper body which meets the flaring lower wall in a sharp junction, and by a single double-handle that curves high above the rim. At some time during the third quarter, a second handle was added to create a distinctive two-handled jug with a mushroom-shaped body—the so-called "mushroom jug"; this vessel is only common in the last quarter of the fifth century and early in the fourth.⁹⁰ As P. E. Corbett remarked: "The introduction of the second handle may of course have been an independent Athenian invention, but published examples from Corinth antedate the earliest known Athenian ones."⁹¹ If Corbett was right in suggesting that a second handle was added to the Athenian shape through the influence of the Corinthian imports, it was only the idea of adding a second handle, not its form, that was adopted, for the double handles of the Athenian vase are quite distinct from the ribbed or flanged Corinthian type. Since the Attic mushroom jug ceases after the early fourth century, it may be that, with the probable dislocation in the Athenian Kerameikos brought about by the difficult last years of the Peloponnesian War and the city's defeat, decanters based upon the Corinthian model and imported from various centers succeeded in ousting the local variety from the marketplace during the first decades of the fourth century.⁹²

Figure 30.

Two-handled oinochoe C-1936-1116 from well 1936-10. Scale 1:3

However, influence went both ways. One two-handled Corinth oinochoe, C-1936-1116 (Fig. 30), from well 1936-10, is rather unusual in shape. It has a very broad flat base, a biconical body with low maximum diameter, two double handles set 90° apart, and a strongly everted rim almost flat on top and horizontally grooved. Given the general shape of the body, and the use of double round handles, one can only agree with Edwards that this is an adaptation of the Attic two-handled "mushroom jug," but with the junction of upper and lower body less sharp, the base flat in conformity with common Corinthian practice, and the mouth modified to produce a more sinuous profile.⁹³ The pale brown clay and the brilliant surface polish enhance the Atticizing impression.⁹⁴ The context would favor a date for the Corinthian decanter in the third quarter, or perhaps early in the last quarter, of the fifth century.⁹⁵ Thus, once the two-handled variety of the Attic mushroom jug had been introduced, perhaps in the third quarter of the fifth century, the new shape soon came to the notice of a potter in Corinth, who tried a local version—although there is no evidence that it had any successors.

Metal Two-Handled Jugs

Metal versions of the two-handled jug do not occur until the fourth century. The examples all come from the area of ancient Macedon. Tomb B at Derveni has provided an elegant jug with goat-head finials at the base of the handles.⁹⁶ However, the high-swung handles and the form of the body are quite unlike the ceramic decanter made in Corinth. Much closer is a bronze oinochoe in Thessaloniki (inv. no. 7438) from a tomb at Stavroupolis.⁹⁷ These metal vases doubtless owe their existence to the popularity of the shape in the second half of the fourth century and to the prestige value of metalware from the reign of Alexander. However, there is no evidence at present to suggest that the shape was produced in metal before the fourth century or that these metal vases exerted any specific influence upon the development of the Corinthian ceramic jugs.

Two-Handled Jugs: Name and Function

In his discussion of the pottery vessels mentioned in the Attic Stelai,⁹⁸ D. A. Amyx conjectured that the term "myke" might have been used as a name for the mushroom-shaped two-handled Athenian decanter of the later fifth century, and many scholars have quite misleadingly latched onto the name to describe all two-handled jugs.⁹⁹ But it should be remembered that, however ingenious Amyx's conjecture, the word "myke" is a restoration, and does not seem to occur elsewhere as the name of a type of vase. In any case, even if the conjecture were to be correct, the word could only be used of the Attic shape, and would not help us to give a name to the two-handled decanter at Corinth. It is worth noting, however, that Amyx also remarked upon the connection in shape between the Classical mushroom jug and the Hellenistic lagynos, and that he considered the possibility that the terms "lagynos" and "myke" might have been synonyms.¹⁰⁰

What about the function of these vessels?¹⁰¹ "The way in which the two handles are set close together may seem strange at first sight; in practice one finds that this setting seems clearly intended to enable one to hand the jug to a reclining or seated neighbor with greater ease and elegance. Furthermore, experiments with an intact specimen have shown that, while the bulk of the contents can be poured out without difficulty, the residue can only be removed when the vessel is turned completely upside down; the dregs are trapped. These two facts suggest that our vase and its ancestors may have served as wine decanters on occasions such as everyday meals, too modest to warrant all the apparatus of a symposium." So, P. E. Corbett, discussing in particular the Attic version of the shape,¹⁰² and he is followed by Edwards.¹⁰³ At least for the two-handled jugs of the Classical period, such as those produced in Corinth and Athens, Corbett's suggestion that the shape was a decanter for wine is surely correct.¹⁰⁴ I can add nothing to his arguments except to note that experiment shows that the squat body of the typical Corinthian two-handled jug in the fourth century would not only retain the lees of wine but also produce a pleasant glugging sound.¹⁰⁵

It is important, however, to give some consideration to changes in shape, particularly in the position of the two handles, between the late seventh century and 146. While handles set 180° or even 90° apart would allow the jug to be passed between two symposiasts sharing a couch, handles only 45–55° apart, as is the norm in the fourth century, would have been impractical for such a purpose. Experience may have shown that, given the capacity of these vessels and the weight of the wine, especially when the jug was full, it was safer to employ two handles placed close together. Perhaps one might also conjecture, from the number of two-handled jugs in the fourth century, that each drinker received his own jug of wine, so that there was no longer any necessity to pass the vessel. There is also the change in the first half of the fourth century to a taller neck and narrower mouth. This was noted by Edwards, who associated the change (from his type II to type III) with the shift, in the second quarter of the fourth century, from the "old broad forms of drinking-cup" to the calyxkantharoi and deep forms with narrower mouth. He surmised that this development required a jug which provided "a more concentrated and more precisely directed flow of wine."¹⁰⁶ Yet experiment shows that types I and II are both suitable for use with the more restricted drinking-vessels of the second half of the fourth century. The Classical series of two-handled Corinth oinochoai comes to an end, as we have noted, by the early third century. The two-handled jugs of the Hellenistic period may have been used not for the consumption of wine, but simply as water pitchers or, at times, even over a fire for boiling and to hold heated contents.

Conclusion

Through a careful examination of archaeological deposits from the American excavations at Ancient Corinth, it has been possible to present, more completely than hitherto, the typological development of the so-called "Corinth oinochoe," both in its one- and two-handled varieties. Both these versions seem to have been introduced during the last quarter of the seventh century. A continuous series of one-handled oinochoai can be traced for some 200 years, well into the third quarter of the fifth century, when the form is replaced by the two-handled version, although some one-handlers may have been produced for another 25 years. The two-handled oinochoe was rarely produced at Corinth during the sixth century, but a clear sequence can be followed from the second quarter of the fifth century. The establishment of such a sequence provides one indicator of gradual change in Corinthian drinking customs after the Persian Wars. During the fourth century, the two-handled oinochoe is the most common form of fineware jug in the Corinthian repertory, and its primary purpose was doubtless for the pouring of wine. However, with the rise of Macedon and the loss of Corinthian political independence, change occurs in sympotic habits at Corinth. By the early third century, the old two-handled wine-jug disappears, and a simpler form, made in fine ware and coarse ware, is introduced, a form that does not seem to have been common, may not have been intended primarily as a wine-jug, and may not have outlasted the third century.¹⁰⁷