the ground. The soft clay being now pressed evenly upon the fabric to the required thickness, the sandy surface of the mould hole easily gave it shape, and gave the potter no anxiety about the outside surface. Indeed, he had but one surface to watch till he came to the in-curve, if his vessel was to have a narrow mouth. Then, I surmise, he built up roughly a clay mould, well sanded, pressing what was left of his fabric into the inside of this mould as he built his vessel upward. Frequently, doubtless, the fabric was not sufficient to go to the top, which explains why sometimes only a part of a jar shows the cord markings.... The distorting and overlapping of the meshes observed by Holmes were probably due to the gathering in to fit the interior of the mould, for it must be borne in mind that the fabric was not shaped in any way to fit the mould, but was doubtless a fragment of some squarely woven article. Thus gathering and overlapping were necessary to make it conform to the inside surface of the mould....” [Illustration: FLUTED VASE, ARKANSAS.] [Illustration: IMPRESSION OF PARTS OF BASKET MOULD ON POTTERY] “When coarse basketry was used for a mould that was intended to be removed before firing, the interstices of the basket work were probably rubbed full of a mixture of sand and clay to prevent the finished vessel from sticking or catching, which explains, I think, the peculiarity of design in some cases, for only the more prominent features of the basket work would impress the vessel.... In some kinds of basketry more filling was necessary than in others, which explains the frequent greater separation and irregularity of the markings.”[90]