in great part only a consequence of the development of art, which rendered it possible to represent a great deal which at an earlier stage could not be depicted. As in the case of male dress, the dresses on the vase pictures gradually become wider with fuller folds. At first, it is true, they are still so narrow that if a long step is taken the shape of the body becomes very distinct; but the cylindrical form, quite free from folds, which the earliest vase style gives to the woman’s dress below the girdle, disappears entirely. Besides this we find, instead of the outer wrap, a puffing, or _kolpos_, which henceforth is the characteristic of the female dress. This was arranged in various ways, though in later times it was sometimes dispensed with altogether. There are different modes of producing this kolpos: sometimes it belongs to the dress itself; the length of the dress then so far exceeds that of the body that, in order to prevent the hem from dragging on the ground, a piece must be drawn up above the girdle, which then falls down in folds below it, and in the fashion of the time, which we have first to deal with, often descends a long way, not, however, equally all round the waist, but only in front, and probably also behind. But as the vase pictures often represent this upper part as of an entirely different material from the under dress, it is possible that it was sometimes not connected with the chiton, but was a distinct garment worn over the under dress, and, like the chiton, fastened in by the girdle. (Compare Fig. 16.) If we remember that in the ancient dress of the previous period, the bib was sometimes a distinct garment, we may surmise that this gradually developed into the kolpos close round the waist, and that the fashion of constructing this girding by means of the chiton itself, and not by a separate piece, was a further stage in this development. With this costume we usually find longish sleeves, reaching below the elbow, as a rule wide and puffed, though very narrow round the armholes. It is evident that a chiton of this description, as well as the upper garment, if it was separate, was entirely constructed by sewing, and was put on over the head by passing the arms into the sleeves; for we nowhere find an opening above the girdle in this dress. We do, however, find, when the upper garment is separate, that the chiton has an opening on one side below the girdle. If we remember the remark of Herodotus previously quoted about the introduction of the stitched Ionic linen chiton, it is a natural assumption that this chiton, which was entirely put together by sewing, and worn without pins, was an Ionic garment; and in accordance with this we find this particular form of sleeve on Athenian reliefs as well as on those of Asia Minor. [Illustration: FIG. 16.]