THE CLIFF VILLAGES OF THE RED ROCK COUNTRY. 571 



examine in vertical section the arrangement of rooms one above another. 

 We excavated a number of almost parallel walls, apparently converg- 

 ing' to the middle of the mound, and found the intervals between them 

 to be cut off by cross walls into series of rooms. The deeper we pene- 

 trated the mound, following the level of the mesa surface, the higher 

 the walls became; and if, as I suspect, these slightly converging walls 

 penetrate to the heart of the mound, there is no doubt that the old 

 part of Awatobi was a compact communal building four stories high. 

 It is likewise evident from the shape of the mound that the western 

 portion of Awatobi was of pyramidal shape, without inclosed court, 

 although possibly penetrated by passageways or narrow streets. 

 Drifting sand and falling walls have combined to form these compact 

 mounds, from which no section of a wall rises above the general level; 

 but wherever we penetrated below the surface there the walls remained 

 unchanged, and if we dug down to the floor through the debris we 

 found the remains of household utensils used by the former occupants. 



The general features of rooms on the south or steep side of the west- 

 ern mound, as shown in the annexed plate, are similar. The masonry 

 of the walls is composed of aligned stones, showing evidences of hav- 

 ing been hewn or dressed with some care at construction. Plastering 

 was generally present, and in many instances was as well preserved as 

 the day the village was deserted. Kafters and floor joists had fallen 

 and been buried in debris, but when dug out were found iu such a good 

 state of preservation that they were available for fuel. The builders 

 who put them in place centuries ago had not known the saw or metallic 

 ax, for each log had been laboriously cut with the stone implements 

 with which he was familiar. 



The appearance of an excavated room of the northern slope of the 

 western mound may be seen in the accompanying plate. Here all indi- 

 cations point to the belief that we have a single-storied room, the raft- 

 ers and other remains of the roof having been removed from the debris 

 excavated from the floor. In one corner there was a small, square 

 closet, and filling the interval between it and the opposite wall a raised 

 seat. The walls were made of small stone blocks with nicely polished 

 plaster, while the floor was paved and covered with adobe. The adjoin- 

 ing rooms over the whole northwestern area of the mounds showed 

 little variation save in details. 



If, therefore, I may be permitted, in the light of my explorations, to 

 reconstruct the original pueblo of Awatobi, as it appeared to the eager 

 eyes of the Spaniards early in the sixteenth century, I should say that 

 it consisted of a large communal building of pyramidal shape, four 

 stories high at the apex, rising abruptly from the edge of a precipice 

 on the south, and sloping gradually northward by a succession of 

 terraces to a range of one-storied rooms. At that time there was 

 probably a spur of rooms extending eastward and facing a dance plaza, 

 in the middle of which was a kiva and shrine. At the close of the 



