100 BUREAU OF AMERICAN ETHNOLOGY [Buy. 185 
from 1.5 to slightly over 2.0 feet. The walls were nearly vertical. 
When newly excavated the rampart of loose earth must have given a 
measurement of nearly 3 feet from its top to the bottom of the trench 
and afforded ample protection to men crouched behind it. Before 
excavation, the bottom of the trench at this point measured from 5 to 
10 inches below the general level, and the outer mound measured 
rather uniformly 5 inches above the level of the surface in front of it, 
although in one place it dropped to but 3 inches in height. 
Features 16 and 17 (figs. 15 and 16).—Two narrow exploratory cuts 
made close together across the fortification ditch were assigned these 
record numbers. These cuts were laid out approximately northeast- 
southwest to cross the ditch at a right angle, and were located a short 
distance southwest of Feature 1, in the northwest part of the site (map 
3). Profiles of these cuts show that the trench at this point had 
sloping sides and that the earth removed from the ditch was thrown 
to both the inner and outer sides of that feature. Before excavation 
the ditch measured, at this point, from 8.5 to 10 feet in width, the 
measurements being taken from crest to crest of the bordering mounds, 
while the depths varied from 10 to 18 inches. The original width of 
the ditch could be accurately determined only for the lower portion, 
the upper part having been destroyed by slumping. In one cut the 
bottom of the original trench measured 2 feet in width, while in the 
other, at a point 20 feet away, it measured 4 feet. The original depth 
of the trench at the points where these exploratory cuts were made 
varied from 18 to 24 inches, which, with the loose soil piled carefully 
in front and behind, probably furnished enough protection for the 
village defenders. 
A post butt 6 inches in diameter was revealed in the center of the 
outer mound by one of these cuts. This butt extended from just below 
the surface of the outer embankment to a depth of 18 inches, its 
base being slightly more than a foot below the original ground level. 
Apparently it was planted before the earth was heaped up to form 
the rampart. Bad weather brought the season to a close before it 
could be determined whether or not an effort had been made to palisade 
the village at this point. Perhaps an occasional post was planted in 
the rampart to carry crosspieces over which robes and blankets were 
thrown to form a screen, a trait reported for the Mandan by Verendrye 
over a hundred years earlier (Haxo, 1941, p. 263). 
CACHES 
Subsurface storage features, indicated by small circular depressions, 
were present but not common at the site, and the depressions were noted 
both inside and outside the ring-mounds. Eixcept for the one found 
and described in connection with Feature 10, none was opened by the 
1951 party. One was tested for depth and character of fill by the 
