The methods of placing the stones in the building have been often debated. The foundations were usually laid on a bed of clean sand, and this enabled the whole course to be accurately adjusted level to begin with. For temples, it seems most likely that the interior was filled with earth as the building advanced; and thus the walls, drums, and architraves could be as easily dealt with as on the lowest course. This plan is successfully used at Karnak in present repairs. But where stones needed to be raised for a pyramid or a pylon, some staging was required. Remains of a massive brick slope still stand against each face of the unfinished pylons at Karnak. This, however, is only the general mass of the staging, and the actual steps for the stones must have been of stone, as brick would crumble to powder if any lifting work was done directly upon it. [Illustration] For short blocks a cradle of wood was used, of which many models have been found in foundation deposits along with model tools. On tilting this to one end, and putting a wedge beneath it, it could be rocked up the slope, and so gradually raised, first to one end and then to the other. For large blocks, the actual lifting was probably done by rocking up. If a beam be supported by two piles near the middle, a small force will tilt it up clear of one pile; on raising that pile the beam may be tilted the other way, and the lower pile raised in its turn. Thus rocking from pile to pile, a beam can be quickly raised till it is high enough to be moved on to the next step. It was probably thus that the fifty-six granite beams, weighing over fifty tons each, were raised in the pyramid of Khufu. The obelisks were transported on great barges, as shown in the sculptures. The method of raising such stones is partly explained by an account of setting up colossi of Ramessu IV. A causeway of earth was made sloping up for a length of a quarter of a mile; it was ninety-five feet wide, and one hundred and three feet high on the slope, probably about sixty or seventy feet vertically, as the slopes were held up steeply with facings of timber and brushwood. The purpose of this evidently was to raise the great block by sliding on its side up the slope, and then to tilt it upright by gravity over the head of the slope. How the mass would be turned we have nothing to show, but probably the simplest way, by gradually removing earth, would be followed. By next ramming earth beneath the obelisk as it lay on a slope, it would be quite practicable to force it forward into an upright position.