The quarrying of the limestone was usually by large galleries run into the best strata. Blocks of two or three feet in size were cut out by picking a trench wide enough for the arm to pass downward around the block, and then inward below it, until it could be cracked away from the bed. The blocks were thus cut out in regular rows, from top to bottom of the gallery face. The same method is still kept up in the open-air quarry at Helwan. For larger blocks a trench eighteen inches wide, in which the workman could pass, was cut around the block. In the sandstone quarries the same mode of cutting was followed, only the quarry was open to the sky. So carefully was inferior stone rejected, that instead of following cracks in the rock, a wall of stone was left on each side of a crack; and such walls, each containing a fissure, divide the quarry to its whole depth. The granite was first obtained from loose water-worn blocks at the Cataract, a great advantage of such a source being that any cracks are made visible. Later it was quarried in the bed; a large mass still in the quarry has been trimmed and marked across to be cut up for shrines or sarcophagi. The early mode of fissuring was by cutting a groove and jumping holes through the thickness of the stone, to determine the direction of the fissure. Probably the active force was dried wood driven in and wetted, as there is no trace of bruising by metal wedges on the sides of the groove. In later times, instead of holes, mere pockets were sunk rather deeper in the groove to hold the splitting agent. For cutting passages or chambers in rock, the method was to make a rough drift-way, then finish a true plane for the roof, next mark an axis upon the roof plane, trim the sides true to the distance from a plumb bob held at the axis, and finally smooth the floor to a uniform distance from the roof. In a rock chamber the roof was finished first, and a shaft was sunk to the intended depth of the chamber to mark it out. The surfaces of rock and of dressed stones were picked smooth by a short adze, with cuts crossing in all directions. The edges of a stone were first dressed true, and then the space between was referred to the edges. To do this, two offset sticks with a string stretched between the tops of them were stood on the edges, and a third offset was used to test the depth to the face, so as to see how much was to be cut away. For larger stones, a diagonal draft-line was cut true as well as the edge drafts, so as to avoid any twist. The face was finally tested with a portable plane smeared with red ochre, and wherever that left a touch of red, the stone was cut down; this was continued until the red touched at intervals of not more than an inch. This was the quality of face for joints; but it was further smoothed by grinding on outer finished surfaces. The rough hewing of rock tombs was generally done with mauls of silicified limestone, which is found as nodules left on the surface.