Another familiar feature is the roll or torus down the corners of the buildings. It is usually ornamented by a pattern of binding. This—as was well pointed out by Professor Conway—is evidently a bundle of reeds bound together, and put down the angle of the plastering in order to preserve it from breaking away. Such a construction was an ugly necessity at first, but when stoneworking arose it had become so familiar that it was faithfully copied in stone as a decoration, and continued to be so copied for more than four thousand years, as long as Egyptian architecture lasted. [Illustration: Fig. 180.] [Illustration: Fig. 179.] [Illustration: 181.—L.D. II. 112.] The well-known Egyptian cornice has been so long taken for granted that it might seem never to have required an origin. Yet in the villages of the Fellahin to-day palm cornices may be seen in course of development. A fence is formed of palm-sticks, placed upright, and stripped of leaves for some way up. The tops are left bushy, and serve to prevent men or animals climbing over the courtyard wall. The upright sticks are tied together by a rope near the top, or lashed on to a cross line of sticks. The fence is stiffened below by interweaving other palm-sticks in both directions; and then the whole is plastered with mud up to the tie level. Here we have the cavetto cornice being formed by the nodding tops of the branches; and to clinch the matter, the earliest representations of that cornice are on figures of buildings which show the crossed sticks of the fence below the cornice. The ribbing of the cornice is seen on the earliest examples, on Menkaura’s sarcophagus in the IVth dynasty (Perring), in the Vth dynasty (L.D. II. 44) and the VIth (L.D. II. 112), and such was copied until late times. But in the more decorative cornices of the XVIIIth dynasty the ribbing was broken up by cross lines, sometimes curved upward, sometimes downward. These cross lines must be a degradation of the leaves of the palm branch. In later times they are omitted, and the pattern becomes simply striped. [Illustration: 182.—L.D. III. 115.] [Illustration: Fig. 183.] This cornice was copied in Syrian architecture, in the plain form without ribbing, as in the tomb at Siloam and the slabs of Lachish; but it does not appear to have ever taken root in Assyria, though attempted there, nor is it known in Europe. [Illustration: Fig. 184.]