64 . BUREAU OF AMERICAN ETHNOLOGY [BULL. 65 
farther downhill or even in the valley bottom. Back of the 
houses come the broken, pinion and cedar covered sandstone slopes. 
There are a few buildings even here, such as the one mentioned by 
Dr. Fewkes on page 11 of his report and cailed by us the “ Tower 
House.” Standing on a spur of rock just below Ruin 8, it was 
once a considerable structure, but is now all fallen except 5 feet 
of the northwestern wall. As in Ruin A, the masonry is double- 
faced, rubble filled, plastered within and neatly smoothed without. 
Some previous digger has cut a hole in the lower portions, showing 
the interior to be choked with earth and stone; a number of slim 
oak and elder twigs visible in the dirt probably once formed part 
of the roofing. 
Near the Tower House and at several other places among the 
ravines are remains of reservoirs, built to supplement the natural 
water pockets that abound in these rocks. To make the reservoirs, 
walls of heavy, rough stones were thrown across little gullies and 
banked with earth. The central parts, probably once 6 or 7 feet 
high, have now been cut away by the water, but the wings, crossing 
the sides of the gullies, can still be made out (pl. 23, a). 
The kiva that was excavated les about half a mile below Ruin A 
and a few feet to the south of its parent ruin (see pl. 1). Although 
part of the southwestern side had been cut away by an arroyo, it 
was otherwise in excellent preservation. It is roughly circular, 10 
feet 6 inches in average diameter, and had evidently been entirely 
subterranean, as the top of its wall, 6 feet 9 inches high, is buried 
a foot or more below the present ground level. 
The masonry is inferior—large, irregular stones set into the 
earth; this was, however, entirely masked with plaster. The. floor, 
lower walls, and parts of the benches still preserve this coating, 
specimens of which show as many as 13 layers. As may be seen 
in the ground plan and elevation of this kiva (fig. 25), it differs 
somewhat from the kivas previously described. The most striking 
features are the large southeastern niche, under which extends the 
horizontal passage of the ventilator, and the long banquette which 
seems to have extended at least three-quarters of the way around the 
room. The two offsets do not quite meet. The horizontal passage 
enters the chamber 7 inches above the floor level; it was roofed with 
cedar slats, now rotted away. The ascending passage is just be- 
hind the rear wall of the deep recess, but is so much broken down 
that no measurements could be taken. The packed adobe floor was 
very firm and smooth; when cleared it was found to be covered with 
half an inch of fine, clean sand, which had presumably sifted in 
after the desertion of the place, but before the roof fell. The de- 
1Kivas were apparently never incorporated directly in the house cluster, but were 
built as near it as convenient; this statement, however, is based merely on surface 
indications. 
