iug away some of the stones to a little depth, they were found to have 

 been well cemented. 



Passing on down the caEiou, not stopping now to notice the more ordi- 

 nary forms of ruins, we passed the mouths of numerous side-caiions, 

 down which came great freshets during the rainy season, gouging out 

 deep arroyos, and strewing the surface with the collected debris of piQDU 

 and cedar, sage-brush and cacti. About the mouth of Coal CaSon, 

 particularly, the whole surface of the " wash " was covered with lumps 

 of fine-looking bituminous coal, as though a thousand coal-carts had 

 traveled that way with their tail-boards out. 



AYe camped at sunset at what our guide called the Rattlesnake Bend, 

 within a half-dozen miles of the outlet of the caiion. Wehad not discovered 

 anymore of the high cliif-houses during the day; but, I doubt not tbat, 

 if we had had a good field-glass with us, many more might have been 

 found along the crevices near the summit of the escarpment. To have 

 verified our suppositions by a personal inspection would have involved 

 a great deal of labor, and more time than we could have spared trom 

 our very scanty store. As it stood, we contented ourselves with secur- 

 ing types. In the vicinity of this camp, the canon changed much in ap- 

 pearance ; instead of the long slope of talus capped by a perpendicular 

 ledge, we have here a perpendicular ledge, first of 200 or 300 feet, and 

 then a long receding bench^ back to the higher mesa beyond. 



Close to our camp was one of the little towers that we now came to 

 look upon as common, about 10 feet in diameter, and remaining some 8 

 feet in height; the inside half-filled with the debris from the walls. 

 Half a mile below, in the vertical face of rock, and at a height of from 50 

 to 100 feet from the trail, were a number of little nest-like habitations. 

 Fig. 5, Plate I, illnstrates their general characteristics. The communication 

 with the outside world was from above, to a small window-like door, not 

 shown in the sketch. Two small apertures furnish a lookout over the 

 valley. The walls were as firm and solid as the rocks beneath which 

 they were built. The stones were more regular in size than any noticed, 

 but smaller. The chinking-in of small chips of stone was noticeably neat 

 and perfect on the inside. This was not a commodious dwelling ; 15 feet 

 would span its length, and 6 its height, while in depth it was not more 

 than 5 feet. Near by, upon a low ledge, and readily accessible from 

 below, was a string of five or six houses, evidently communicating, mere 

 kennels compared with some others, made by walling up the deep cave- 

 like crevices in the sandstone. The same hands built tbem that lived 

 in the better houses; the masonry being very similar, especially the 

 inside chinking, which was perfect, and gave the walls a very neat ap- 

 pearance. Fig. 8 of Plate II is an example of the tenacity of their mortar; 

 the view being of one of the line of little houses just spoken of. In this 

 case, a portion of the ledge upon which the house stood has become 

 separated from the cliff, carrying a portion of one of the buildings with 

 it; and although torn away from the remaining wall, and thrown over at 

 a considerable angle, yet it remains perfectly firm and unshaken. 



Scratched into the face of the cliff which contained these houses were 

 various inscriptions, one of which I have depicted in Fig. 6 of Plate I. 

 As they are not cut in very deeply, and in some places mere scratches, 

 I doubt very much whether they are contemporaneous with the houses 

 themselves. 



Two or three miles farther, and the caiion changed in feature again ; 

 the highest level of the mesa coming forward and towering over the 

 valley with a thousand feet of altitude ; the bottom-lands widening out 

 to a half and three-quarters of a mile in breadth. Cottonwood and 



