KIDDER-GUERNSEY] ARCHEOLOGICAL EXPLORATIONS IN ARIZONA 59 
courses of the wall.t. At some time after the completion of this room 
it so settled into the ground that the tops of the walls sagged away 
an inch or so from the cave roof; the space thus formed was stopped 
up all around with adobe of a different color from that used in the 
original masonry. 
Although nowhere in Ruin 8 were any roofs in place, fragments of 
roofing were found in the débris. Over the beams had been laid 
twigs or cedar bark,? and upon this was spread a coat of adobe 
mortar, over which were laid bunches of long grass, more adobe, 
more grass, and finally a last layer of adobe, making a total thick- 
ness of about 4 inches. We have not observed this grass-layer 
Q6IGse 
Bex 
ZSerae 
@ 
e@ 
Fie. 24.—Kivas 1 and 2, Ruin 8. 
method in any other ruin, nor have we seen it described. It makes 
a very strong and springy covering, and pieces of the mixture will 
stand a blow which would shiver an equal amount of unmixed 
adobe. 
Of the two chambers on the eastern front, K 1 is probably, and 
K 2 is surely, a kiva (fig. 24). Both are in poor preservation, water 
from the cliffs above having so entirely destroyed their front walls 
that their ventilating shafts, if they had them, have quite disap- 
peared. Both had once occupied a semisubterranean position, with 
their rear or northern portions well sak in the cave deposit, their 
southern walls probably standing partly free. 
1For the designs see fig. 97, a, b, oc. 
2We saw neither reeds nor slats. The former were found in Ruin 2, and the latter 
observed in place in the roof of a cliff-house in Devil’s Canyon, San Juan County, Utah 
(see Kidder, 1910). 
