90 BUREAU OF AMERICAN ETHNOLOGY [BuLu. 185 
at the Dodd and Phillips Ranch sites the pit walls are described 
as made up of “unfaced refuse and native soil.” Unfortunately, one 
does not know whether this indicates a mixture or whether the pit was 
dug through a layer of refuse into native soil. It may well be that 
at these two sites the depth of the pit is made up largely of an 
accumulation of refuse which grew up about the house walls and over 
the entire village area during the occupation of the site. Certainly 
this is the impression one gets at the 89ST1 site, which is probably 
of approximately the same age as the Dodd and Phillips Ranch sites. 
Vertically framed walls with no traces of leaners and with walls 
outlined by many closely set posts in a pit, which may be rather deep 
or very shallow, are typical of the houses of the Upper Republican 
and Nebraska aspects in Nebraska and Kansas. The later round 
structures excavated from Lower Loup sites show, generally, an outer 
circle of 16 posts and, although no traces of leaner butts have been 
found, some such arrangement must have been present. That such 
was not always the case is indicated by one of the few floor plans 
which have been published; House 1 at the Gray-Wolfe site showed 
a circle of 112 small posts spaced at intervals of 10 to 20 inches at the 
edge of the floor (Dunlevy and Bell, 1936, pp. 164-165). A house 
thus outlined would not require leaners for the support of the earthen 
walls. Since there is no trace of leaners for most of the earlier struc- 
tures, their use may be a late trait, and rather earlier in the south than 
in the north. However, small, out-of-pattern houses may have used 
them at an earlier period as I have suggested in another place (Met- 
calf, 1962). It is possible, of course, that the Mandan-Hidatsa group 
may have used leaners earlier than did the Arikara and that the 
trait was borrowed from them by the latter group, along with the 
Mandan-Hidatsa type of entranceway. At present it would seem that 
an amount of study comparable to that given to the pottery from 
a site must be given to the floor plans of the structures and the village 
arrangement before a true picture will be obtained for the cultural 
changes on the Plains. 
OTHER STRUCTURAL REMAINS 
The remains of three structural features other than earthlodges 
were uncovered at this site. It was impossible to determine with 
certainty the nature and use of these structures. 
Feature 10 (fig. 11) —Immediately adjoining Feature 8 on the west 
was a small earthwork similar to those marking house sites, but rec- 
tangular in outline. It was not oriented to the cardinal directions, 
and Stout in 1908 indicated an entrance at the southwest end. There 
was no surface evidence for an entrance in 1951, either as mounds 
resulting from the collapse of an earth-covered and earth-banked 
vestibule or as a break in the continuity of the earthwork. Stout’s 
