If still further proofs be required of the imitative character of this early stone architecture, we shall find them in the door of a tomb (Fig. 35). Nothing can be clearer than the way in which the lintel obtained its peculiar character. It is formed of a thick slab engaged at each end in the upright beams of stone which form the jambs. This slab appears beyond the jambs, and ends in a deep groove, which divides them from the walls. Underneath the lintel, and well within the shadow which it casts, there is another and more curious slab; it is, in shape, a thick cylinder, corresponding in length to the width of the door. In the deep groove already mentioned the ends of the spindles or trunnions upon which it is supported are suggested. They are not, indeed, in their right places: they are too near the face of the building. The workman would have had to make the groove very deep in order to show them in their proper places, and he was therefore content to hint at them with sufficient clearness to enable those who saw them to understand what they meant. We have none of the wooden models under our eyes which were familiar to the stonemason who carved these doors, but yet we can easily see the origin of the forms we have just described. The cylinder was a circular beam of acacia or palm, upon which a mat or strip of cloth of some kind was nailed. By means of coils in the groove at the side the cylinder could be made to revolve, and the curtain would thus be easily drawn up and down. These curious forms are thus at once accounted for if we refer them to the wooden structures which were once plentiful but have now disappeared. Nothing could be more difficult than to find an explanation of them in forms appropriate to stone or granite. Of what use could such a cylinder be if carried out in either of those materials? It could not revolve, and the deep lateral grooves, which have such an obvious use in a wooden building, would be purposeless. [Illustration: FIG. 35.--Door of a tomb at Sakkarah; drawn by Bourgoin.]