96 BUREAU OF AMERICAN ETHNOLOGY [Buun. 185 
ness to the hearth. Some such structure would account for Features 10 
and 15 as well as any explanation thus far advanced. If such chim- 
neys were present above the fireplaces one would expect a certain 
amount of lumps of burned earth to have been present in and about the 
hearths. Such evidence was not present. Stout’s informant said 
Feature 10 had served as a stable. It may well have also served as 
a shelter for men. 
Feature 13 (fig. 13). —Before excavation Feature 13 showed on the 
surface as four very low mounds forming a rectangular enclosure. 
These were much lower than the ring-mounds of house sites and even 
lower than those at Features 10 and 15. The feature was located 
southwest of the village, across the ditch and between that feature and 
the county road (map 3). Locally it was believed to mark the site 
of a cabin occupied by a trader who came to the site with the Arikara, 
although in 1908 Stout was told that the trader occupied an earthlodge 
in the village. Unfortunately, Stout obtained no information regard- 
ing Feature 13. 
In excavating this feature, an area measuring 50 feet east-west and 
40 feet north-south was laid out to enclose the mounds completely, 
after which all soil was removed to a depth of 6 inches below the 
surrounding level surface. When the floor of this excavation was 
cleaned and smoothed, a faint discoloration of the soil could be seen. 
This could not be accurately delimited since, faint at best, it merged 
without a sharp break into the surrounding soil at a point below where 
the mounds stood. The excavation was then deepened by scraping 
with trowels and shovels until it was lowered another 2 inches. At 
this point all discoloration had disappeared, and nearly 170 small 
molds from 1 to 2 inches in diameter were revealed. These were but 
a few inches in depth and not infrequently contained a small bit of 
decayed wood, presumably of willow or other saplings. These rods 
had been planted, not in rows, but irregularly in a band from 22 to 
36 inches wide and measuring 33 feet in length east-west, 32 feet 
in width at the east end, and 27 feet at the west end. The mold-free 
area thus enclosed measured approximately 25 to 27 feet in each direc- 
tion. Noartifacts came from the excavation. 
I find it impossible to determine what, other than a brush-fenced 
enclosure, is represented here. It definitely was not a log cabin nor 
are any of the molds large enough to have held posts of the size 
required for a ramada or a drying platform. In some ways it is 
reminiscent of a brush-fenced tobacco garden described for the Hi- 
datsa (Wilson, 1917, pp. 126-127), and it may have been such. On the 
other hand it may have served as a small corral in which horses were 
penned. The low mounds about it may have resulted from blowing 
soil trapped by weeds caught and held by the closely set rods of the 
fence, although this seems highly improbable. 
