It is to all intents and purposes a stone house. Two kinds of rocks predominate among the material; a slaty, gray and red, sandstone,--highly tabular, easily broken into plates of any size,--and a sandstone conglomerate, containing small pebbles from the size of a pea up to that of a small hazel-nut,--the whole rock of a gray color. When freshly broken or wetted, this conglomerate becomes very friable, and so soft that goats have left the impression of their feet on scattered fragments. When dry it becomes hard, and is always very heavy. Both kind of rocks are found in the vicinity of the _mesilla_. Besides these, loose pieces of stone from the bluff itself, boulders from the creek, of convenient size, enter into the composition of the walls. Sometimes the latter consist exclusively of slabs of sandstone superposed; again there are polygonal fragments of rocks piled upon one another, with courses of tabular sandstone, forming, so to say, the basis for further piling; the foundations are usually boulders and the hardest rocks, also of greater width. There are no walls of dressed stone, but the rocks are broken to a suitable size, as may be done with any stone maul or sledge, or even by smashing with the hand and another rock. In fact the whole stone-work must be termed, not masonry, but simply judicious and careful piling.[107] In performing it, great attention has been paid to having the vertical surfaces as nearly as possible vertical; but this end could be reached without the use of the plumb-line, and with the aid of mere ordinary eyesight, for the rooms are so small, and the partitions so thin, that anything not "true" could, and can yet be, "shoved" into position by a mere steady, slow push; carefully watched on the opposite side. The same applies to the angles, although they are tolerably accurate. As a general thing, the transverse walls appear to be continuous, and the longitudinal partitions to have been added afterwards, but there are also instances of the contrary. In this respect the sinuosities of the rocky foundation seem to have determined the mode of action. To fill up the gaps between the stones, and to coat them with a smooth surface within the chambers what appears to be earth from the surrounding bottoms has been flung into the crevices, thus forming a natural mortar, and at the same time a "first coat" of plaster of varying thickness. This in turn is covered with a thin white layer (now of course turning into gray, yellow, and flesh-red) much resembling our plaster, but whose composition I am unable to determine. (Specimens of the mud, containing small gravel and minute particles of mica, are sent with the other collections, also fragments of the white coating for analysis.[108])