44 BUREAU OF AMERICAN ETHNOLOGY [Buny. 185 
The cultural affiliation of this site is unknown at present. A vague 
and general relationship to Mandan and Hidatsa material is present 
in the artifacts, but the texture and tempering of the sherds differ 
markedly from sherds recovered from 32ML39, from Rock Village, 
and from the sites of the Hidatsa villages at the mouth of the Knife 
River. The majority of the rim forms are either rare or not present 
at 82ML39 and Rock Village, although a minority group closely re- 
sembles the bulk of the ware from those two sites. The relationship to 
the culture represented at the Hagen site is no closer than that to the 
Mandan-Hidatsa sites. 
An unpublished manuscript report by Hecker, 1936, in the files of 
the State Historical Society of North Dakota contains references to 
three sites in Saskatchewan which appear to be more closely related 
to 82W112 than any others on record. The first of these sites consists 
of a midden on a coulee slope. Many stone circles were reported to 
have been present on the adjacent upland before the level land was 
placed under cultivation. The second site, which has been extensively 
vandalized, is at the junction of Moose Jaw Creek with the Qu’Appelle 
River. The midden deposit there is described as being from 10.0 cm. 
to 1 m. in thickness and covered with 15.0 to 25.0 em. of soil. The 
third site is about 2.5 miles east of Last Mountain Lake. There again 
the site consists of a midden on a coulee slope. The coulee is deep, 
narrow, and heavily wooded, and Hecker believed that a camp had 
been present in the bottom of the coulee—from the location, probably 
a winter camp. 
No contact material is reported from these sites, a knife blade from 
the last being considered to be of native copper. No pottery is men- 
tioned from the last site, and that from the others is described in rather 
general terms, but all traits given for it are present in the ware from 
32W112. Stone tools from the sites include end scrapers, drills, small 
blades, and small, side-notched projectile points. Side-bladed rib- 
knife handles, flakers, ornaments, rib-bone end-scraper handles, and a 
few miscellaneous pieces of indeterminate use make up the inventory 
of bone objects from the sites. 
In these three Canadian and one North Dakota sites we appear to 
have the components of a new complex. It is to be hoped that con- 
trolled excavations at them can be made before the existing evidence 
is destroyed by relic hunters. 
At present there is no evidence to indicate that the material re- 
covered from the midden in the coulee at 832WI12 is related in any 
way to the tipi rings on the ridge above them. Comparative material 
from the stone circle site (and similar sites) is lacking. It must be 
pointed out, however, that there are few places in the coulee for a 
village site, and no other traces of habitation remains were found 
other than the rings. On the other hand, it is by no means certain 
