Many theories have been advanced as to the various garments which compose the costume. It will be well to give a brief summary of them, and to point out wherein they fall short, and, if possible, to substitute one that is more satisfactory. The chief point at issue is whether the skirt part of the drapery belongs to the chiton—that is to say, to the garment which appears on the neck and left arm—or whether it is part of the cloak which passes under the left arm and is fastened on the right shoulder. Collignon even distinguishes three garments; he believes that the skirt is the chiton proper, and that the crinkled texture of the piece which appears above the himation is meant to represent some sort of woollen jersey worn over the chiton, which he calls the “chitoniscus.” The difference in texture comes out very plainly in those cases where the himation is worn over the shoulders like a shawl, or where it is omitted altogether; for example, in Nos. 670 and 671.[128] [128] Lechat, figs. 8 and 9; Perrot and Chipiez, 290 and 292. At first sight it appears as though two separate garments were intended, but on close examination it will be found that the curved line which terminates the wavy lines of the upper section has not the appearance of an edge, but appears rather to turn under and to represent a pouch, formed by pulling the garment up through the girdle. Moreover, in some cases these parallel wavy lines appear on the skirt as well, and cover the whole surface with the exception of the mass of folds hanging down the middle of the front. This can clearly be seen in No. 687 (Lechat, p. 161), in a small statue of the same type from Eleusis, now in the National Museum, Athens, and in the relief of the Charites in the Acropolis Museum (Lechat, pl. 3). Again, the same technique is found sometimes introduced into the rendering of the himation. Frequently on the shoulder, when the cloak is fastened, a succession of these wavy parallel lines begins to appear, then stops suddenly, and the rest of the garment presents a smooth surface.[129] There can be no question here of a difference of material, nor of a separate piece of drapery, so that we must look for some other explanation of the different treatment. Lechat has offered one which is satisfactory and which finds confirmation in other monuments. He says “the difference in the appearance of the upper and lower part of the same garment is due to this: that in the lower part, all the superfluous material is gathered together in a single mass, and the rest is drawn tightly across the legs; while in the upper part, the material, being left free, falls in regular folds all round the body.” He further suggests that the regularity of the folds may be meant to represent some artificial treatment of the dress, such as is applied to the modern _fustanella_. The archaism of the work, however, is sufficient to account for this regularity in representing a series of very full folds in a fine material held