Pottery has been the chief instrument, too, in the formulation of a system of dating. By assuming a lapse of a thousand years for every yard of deposit--except in the Stone Age, when the accumulation of debris was quicker, because huts were built of ephemeral material such as mud and wickerwork--each successive layer is relatively dated according to its depth from the surface. Pots provide the nucleus for this scheme, being found in large numbers in every layer. Other objects take their place according to the type of pots they are found with. Not that the process is simple. There are complicating factors, and even pottery creates difficulties and irregularities. At Knossos, for instance, when the first palace was built, the top of the hill was levelled and a portion of the former deposit thus cut away. Obviously, too, heirlooms would belong to an earlier time than that of the layer in which they are found. Or a pot may be displaced in the earth. A safeguard, however, against mistakes is afforded by the abundance of pots, which makes the differentiation of general classes easy. Pots, then, are found at the lowest levels, just above virgin soil, for the earliest people used them and broke them. The slowness of development in that long-drawn-out period (the Neolithic or Later Stone Age) is clearly indicated. There are some seven yards of deposit belonging to it at Knossos, and the latest ware shows little or no improvement on the first. The pottery is hand-made, the clay coarse, generally of a sooty-greyish colour and more or less burnished. The relics consist of the rims and handles of pots, rims of basins, bowls, and plates and similar fragments, too incomplete to suggest original shapes. Two interesting points, however, can be seen. The pots were hand-polished both inside and out, and incised lines, or lines simply scratched on the surface, were used as ornamentation. This primitive manifestation of an artistic impulse was later extended by the filling of the incised lines with a white substance for greater effect. Similar ware has been found at Troy and in Egypt, and Dr. Mackenzie has thought that these were an importation from the Ægean (_Journal of Hellenic Studies_, vol. xxiii, p. 159). The irresistible impulse manifested even in primitive people to decorate their ordinary vessels is further illustrated by the fact that the polishing was gradually heightened, and the glitter thrown into relief by ripples, made with a blunt instrument, probably bone, and suggestive of the ripples on the surface of water. Among the latest Neolithic ware found at Knossos are two remarkable specimens of incised ware, the design being that of a twig with leaves. On each side of the stem is a row of small oblong punctuated points, filled in with white chalk. This, it must be remembered, in a period which ended about 3000 B.C.