58 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1958 
work there would provide information on such lesser establishments 
of the fur and Indian trade, of which little was ever recorded at the 
time they were in use. Excavations at 391.M241 proved that it was 
not the site of Fort Defiance (or Bouis) but that it was of a later 
period of permanent settlement, dating after 1880. Further search 
for the Fort Defiance site proved fruitless. While somewhat scanty, 
the data and specimens from 39L.M241 provide materials that should 
be very useful for comparative studies relating to this later period of 
white occupation. The Smith party completed 6 weeks of fieldwork 
and returned to the Lincoln office early in August. 
The second River Basin Surveys field party in the Big Bend Reser- 
voir area at the beginning of the year was directed by Dr. Warren W. 
Caldwell and consisted of a crew of nine. The group was at work on 
the right bank of the Missouri River in Lyman County, S. Dak., 
some 7 miles above the Lower Brule Agency, excavating in the Black 
Partizan site (39L.M218). The latter consists of the remains of a pre- 
historic earth-lodge village of at least 2 component occupations and 
perhaps 8. The party completely excavated 1 circular earth-lodge 
ruin and a large portion of a second, cross-sectioned a defensive forti- 
fication ditch, excavated 1 complete bastion of the stockade, and tested 
a number of midden areas and cache pits. The circular houses were 
situated well outside the fortification ditch and were of the late occu- 
pation of about the end of the 17th century. The ditch and bastion 
represent two earlier occupation periods, with the ditch being dug 
during the earlier one, later filled in and, still later, redug. Tests 
indicated a rectangular house inside the fortification ditch and, to- 
gether with midden areas and cache pits in that area, provided both 
simple-stamped and cord-roughened pottery that predates the mate- 
rial recovered from the circular houses outside the stockade. Among 
the finds made in the cache pits, one of particular interest was the 
burial of two very large, adult dogs, together with a pup. After 12 
weeks of excavation, the party disbanded and returned to the Lincoln 
office on September 7. The Caldwell and Smith parties shared a 
joint field camp near the mouth of Medicine Creek. 
The third River Basin Surveys field party in the Big Bend Reser- 
voir area at the beginning of the year was under the direction of 
Robert W. Neuman and had a crew of 10. That party conducted 
excavations in four sites in the vicinity of Old Fort Thompson, the 
Indian Agency, on the left bank of the Missouri River in Buffalo 
County, S. Dak. Two of them were prehistoric village sites on the low 
terrace bottoms, and two were burial mounds situated on the higher 
terrace of the Missouri River. The Pretty Bull site (39BF12) was 
found to have had three separate occupations. The earliest and 
deepest remains were recovered from two test excavations that un- 
