BRACED-UP CLIFF, PUEBLO BONITO—JUDD 507 
wide at the bottom and 6 feet wide on top, 7 feet high, and at least 80 
feet long. 
But the Bonitians were not content with this achievement. Before, 
or perhaps after, completion of that embankment they undertook to 
support the undercut cliff directly by adding wooden props, rubble- 
work, and more adobe mud, concealing all behind a buttresslike con- 
struction I shall describe as the “upper stonework,” in order to dis- 
tinguish it from the terrace and the terrace wall. 
This upper stonework likewise consisted of a masonry veneer screen- 
ing the mud and broken rock massed behind. But here the purpose of 
the hidden rubblework clearly was to replace the softer sandstone that 
had weathered away and thereby created an irregular, cavelike recess 
varying in depth from 0 to 15 feet. The rubblework behind the 
veneer had filled every nook and hollow of the cave and crowded its 
roof, as evidenced by mud mortar still adhering. Unlike that of the 
lower terrace, however, the facing of this upper stonework is non- 
descript and utterly without character; its finished appearance is 
due solely to the fact that its component fragments were positioned 
with their wider edges to the fore. Built as a wall perhaps 5 feet 
thick at the base, standing fairly erect at the back but with a pro- 
nounced cliffward batter in front, this facimg masonry reached to the 
cleavage plane that marks the top of the underlying softer stratum 
and there, reduced to a probable thickness of 2 feet, was fitted snugly 
against the overhang. Minimizing labor, the builders erected their 
buttressing stonework only where the undercutting was deepest, a 
30-foot section at the east end of the cliff and one twice as long at the 
west. The height approximated 15 feet (pl. 3, fig. 2). 
Behind the sloping exterior of the upper stonework are a number of 
casual partitions dividing the rubble fill and a dozen pine props, the 
feature that gave the Braced-up Cliff its Navaho name. Spaced ir- 
regularly, these props vary in diameter from 10 to 12 inches. We 
counted nine, one of them at the extreme rear of the cavity, and the 
empty holes of four more (fig. 2). All but the two most inaccessible 
had been cut off as low as possible with steel axes or had been burned 
long ago and subsequently weathered (pl. 3, fig. 1). Each post 
slanted rearward to meet the roof of the cave; each had been tightly 
packed about with rubblework. We sectioned Nos. 2 and 7 (JPB 
156, 157) which later were dated by Dr. A. FE. Douglass at A.D. 1057 
and 1004. Douglass (1935) and Smiley (1951) both list these two 
specimens in their tables of tree-ring dates but do not specifically 
identify them. For Gila Pueblo in 1940, Deric O’Bryan sampled 
posts 2, 3, 4, 5(%), and 7 and obtained cutting dates of, respectively, 
1058, 1072, 1064, 1061, and 1073 (personal communication). The 15- 
year spread in these latter readings and a date 54 years earlier for 
Douglass’ No. 7 suggests that each prop may have been the leftover 
