176 THE CLIFF RUINS OF CANYON DE CHELLY  [erm.any.16 
wide, and the ends of the eurved walls terminate in smoothly finished 
surfaces. In front of it there are remains of two walls, about a foot 
apart, and so arranged as to form an apparent passageway into the inte- 
rior of the kiva. These seem to be a kind of platform, like that just 
described, but close inspection shows the walls, which can be traced 
to within 6 inches of the inner wall of the kiva. This also may be the 
remains of a chimney-like structure. There are other points in the can- 
yon where the same feature occurs, but in none of them is the evidence 
of an opening or doorway more definite than in the examples described. 
The masonry of the kivas is always as good as that of any other strue- 
ture on the site, and generally much better. The walls are usually mass- 
ive; sometimes they are 3 feet thick in the upper part and 4 feet in the 
lower portion, where the bench occurs. In a few cases the kiva has an 
upper or second story, but when this occurs no attempt 1s made to pre- 
serve the circular form, and the upper rooms are really rectangular with 
much rounded corners. Plate XLIx shows a second-story kiva wall in 
Mummy Cave ruin, and plate Lxmr one in ruin No. 10 in De Chelly. 
The latter occurs over the prin- 
cipal kiva, and the walls which 
are still standing on the north 
and west sides are approximately 
sm straight, but the corners aremuch 
rounded. Figure71 is a detailed 
plan of part of the kiva, show- 
ing the arrangement of the upper 
walls. The kiva walls are about 
18 inches thick. On the north 
side the upper wall is supported 
by a heavy beam, part of which 
is stillin place. Under the north- 
Big, 71 Elaworpartct a }iva tnzuin Note) -eastecomen of ithesuppertcoom 
there is a little triangular space formed by a short connecting wall, 
shown on the plan. This is really a flying wall, covering only the 
upper portion of the space, and its purpose is not clear, as the open- 
ing left is not large enough to permit the passage of a person, and was 
available only from the second story. 
Apparently the greatest care was bestowed on the construction and 
finish of the kivas. The exterior of the circular wall is often rough 
and unfinished, but this is probably because the whole structure was 
generally inclosed within rectangular walls. The interior was plastered, 
often with a number of coats. The southern kiva in ruin No. 10 shows 
a number of these on its interior surface, applied one after another, and 
now forming a plastering nearly three-quarters of an inch thick. In its 
section 18 distinct coats can be counted, separated one from the other 
by a thin film of smoke-blackened surface. The kiva in ruin No. 16 
has 4 or 5 coats, that in ruin No. 31 shows at least 8. In the last 
example the last coat was not decorated, but some of the underlying 
3 
ee | 
fait 
Larsxg 
Shits 

ones were. 
