There remains the “certain part of a furnace” referred to by Hesychius. Whatever this may have been, it is most unlikely that it was rectangular for the simple reason that ancient Greek furnaces appear from extant pictures to have no rectangular parts that could possibly be so identified. Pictures of ancient Greek furnaces are numerous, and it is surprising that no attempt seems to have been made to discover a cypsele in any of them, for it is not unlikely that some of these pictures do in fact depict it, and in that case they show us the earliest form of the object of which we have any precise record. [Illustration: Fig. 29. Attic vase painting, perhaps depicting a cypsele.] [Sidenote: Probable pictures of a cypsele on sixth-century Attic vases, depicting it as a large terra cotta vase.] In Saglio’s _Dictionnaire des Antiquités_, s.v. Fer, figs. 2964, 2965, (here figs. 27, 28), pictures are reproduced from two black figure vases that depict furnaces being used, in all probability for treating iron[1051]. On the top of either furnace is depicted what de Launay describes as a “sorte de vase, sans doute de terre cuite[1052].” A similar vase is shown _ibid._ s.v. Caelatura, fig. 937[1053] (here fig. 29) on top of what the context shows to be the furnace of a bronze foundry[1054]. This latter picture is on an early red figure vase[1055]. In all three cases the “sort of vase” is very large, as is shown by comparing it in size with the human figures in the picture. In short both in size, shape, use and probably material it answers to the written descriptions of one variety of cypsele. What it does not so well answer to is the vase represented on the coins of Cypsela, which is tall and cylindrical and shows two vertical handles. Fortunately there is a connecting link between the two forms. In the Berlin Museum there is an actual stove of terra cotta, said to have been found in the sea off Iasos (coast of Asia Minor)[1056]. It is about ·50 m. high and of a common enough type[1057], though it is rare to find one so well preserved[1058]. What however gives the Iasos example its importance is a vase ·13 m. high and ·192 diameter, of the same dark brown micaceous clay as the stove itself, that was found along with it and fits so well on top of it that it must unquestionably have formed part of the complete article[1059]. Here we have a vase of considerable size that in shape has resemblances to the vases on the coins of Cypsela, but in its bulging sides deviates from them in the direction of our conjectured cypselae of the vase paintings. In position and in use as the receptacle for material to be heated, it corresponds with these latter. Though comparatively large, it must remain doubtful whether it is large enough to be a cypsele. But in any case it helps to connect the vases of the Cypsela coins with the “part of a furnace” of the vase paintings, and to make it probable that the latter was made of pottery and that both are