MINDELEFF] THE ABSENCE OF BUTTRESSES 107/33 
wherever horizontal timbers are used for the support of masonry they 
rest on the bed rock. 
The same ruin (No. 32) contains an elaborate system of retaining 
walls, which are shown partly in figure 69. At first a retaining wall 
was built immediately in front of the main kiva, which is now 5 feet 
high outside. Apparently this did not serve the purpose intended, 
for another and much heavier wall was built immediately next to it. 
This wall is 4 feet thick, flush on top and inside, but 10 feet high out- 
side. At halfits height it has a step back of 6 inches. It would seem 
that even this heavy construction did not suffice, and still another wall 
was built outside of and next toit. This wall is nearly or quite as heavy 
as the one described, and its top is on the level of the foot of that wall, 
but it is 12 feet high outside. Something of the character of the site 
may be inferred from the arrangement of these walls, which have a 
combined vertical fall of 27 feet in a horizontal distance of less than 15 
feet. The outer or lower wall has a series of very heavy timbers pro- 
jecting from its face; these are placed irregularly. It should be noted 
that access to this village was from the bench on either side, and that 
it could not be reached from the front, where these walls occur. There 
are other walls on the lower slope, similarly reinforced. 
A little to the right of the point where these retaining walls oceur 
there is a room in which horizontal beams have been incorporated in 
the masonry. A similar use of timber occurs in ruin No. 16 and is shown 
in plate Lx. Why timber should be used in this way is not clear. It 
may be that when the supply was placed on the ground the builders 
found that they had more timber than was needed for a roof and used 
the excess in the wall rather than bring up more stone. The posts 
which were placed vertically and built into the wall were always short; 
perhaps they were fragments or ends cut from roofing timbers that were 
found to be too long. In many instances they failed to hold the walls, 
and possibly the pit holes in sloping rock, which are numerous on some 
sites, indicate places where this expedient was formerly employed. 
It is singular that the necessity for such expedients did not develop 
the idea of a buttress. On this site such an expedient would have 
saved an immense amount of work. In only one place in the canyon 
was a buttress found. This was in the Casa Blanca ruin, shown in 
plate xitvir. There is no doubt that in this place the buttress was 
used with a full knowledge of its principles, and but little doubt that 
the idea was imported at a late, perhaps the latest, period in the oceu- 
pancy of that site. Had it been known before, it would have been used 
in other places where there was great need for it, net so much to pre- 
vent the slipping of walls as to supersede the construction of walls 4 
feet thick or more, and to strengthen outside walls which were likely to 
give way at any time from the outward thrust upon them, : 
Altogether the constructive expedients employed in De Chelly sug- 
gest the introduction of plans and methods adapted to other regions 
