34 BUREAU OF AMERICAN ETHNOLOGY [ BULL. 65 
3 to 4 feet high, nor were they ever, apparently, much higher; at the 
north end they are much broken down. There were no doors in the 
series of rough-walled rooms. These doorless inclosures were seem- 
ingly never roofed over, and were probably outdoor living rooms, 
used in preference to the open cave space in order to have a level 
floor and to afford protection from the wind. 
The two bits of wall on the upper level at the rear were attempts 
at terracing abandoned before completion. Behind the southern one 
there is an empty metate bin of stone slabs. 
The only features of architectural interest are furnished by the 
more carefully built tier of four rooms. The first three of these 
are large enough for habitation; the fourth, a little closetlike cham- 
ber, was used presumably for storage purposes. Each of the first 
Fie. 11.—Plan of Ruin 3. 
three has two doorways opening upon the narrow passageway be- 
tween the rooms and the rock ledge behind; in room 1 and room 3 
one of the two doors is sealed up with masonry; in room 2 both are 
open. The purpose of so many entrances is rather puzzling, but may 
perhaps be explained by the presence in room 2 of traces of an 
adobe wall which once divided it into two small square apartments. 
This wall had evidently been removed before the house was aban- 
doned, but both doors were left open. In the other two rooms there 
were probably also such partitions cutting the chambers in two and 
necessitating double entrances; when the partitions were taken out 
one of the useless apertures was closed. The doorways them- 
selves are in excellent preservation, differing from each other only 
in very minor details of measurement, and are typical of the doors 
