These statements quite suit the list of uses. As between the two materials mentioned the Aristophanes scholiast gives the impression that the commoner was wicker or basket work. But in the Lucian scholiast wickerwork is only an afterthought added by a later hand. The Lucian scholiast is probably more correct. Pottery is a natural material for every kind of cypsele. Cypselae (1) and (3) can never have been of basketwork, and for (5) it looks a most unlikely material, though we know too little about ancient spice vessels to speak with certainty. For (2) it is suitable enough, but the cypselae of the Thracian coins, which the emblem of the grain of corn shows to have been probably corn jars, point in the other direction. Their shape suits either terra cotta or metal but not wickerwork. The probable use practically excludes metal[1043]. (4) is according to Hesychius a “plaited beehive,” _i.e._ a hive of basketwork, and this statement is accepted by M. Pottier[1044]. No doubt it was true at the time when it was made. But is it so certain that the earliest cypsele beehives were of basketwork? The first reference to beehives is in the _Odyssey_, which describes hives of stone in the shape of vases (κρητῆρες and ἀμφιφορῆες)[1045]. These Homeric beehives must have been either prototypes or glorifications of hives of earthenware[1046], and it is tempting to classify these latter with the cypsele form of hive, especially in the light of the cypsele on the Thracian coins, which has much the shape of a mixing bowl except that the handles are of a type more frequent on the amphora. These coins are of the first half of the fourth century. Our literary authorities are all much later. Most of them mention earthenware beehives only to condemn them[1047]. Presumably they were out of fashion. Basketwork hives[1048] on the other hand are spoken of without condemnation. When therefore they define the cypsele beehive as a basketwork beehive, they practically mean a round or vase-shaped hive like the “little pail where the bee distils his sweet flow” of Antiphilus[1049], as distinguished from the rectangular form that was also much in use[1050]. This is assuming that cypselae were never rectangular, but the assumption seems fairly safe. Neither plaiting nor pottery adapts itself to rectangular shapes. Wine vessels are not usually square. The cypsele of the coins is not rectangular. [Illustration: Fig. 27. Attic vase painting, perhaps depicting a cypsele.] [Illustration: Fig. 28. Attic vase painting, perhaps depicting a cypsele.]