34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I26 



becomes a sizable stream with a relatively uniform flow in a valley 

 which supports rather heavy stands of timber. There is a correspond- 

 ing variation in the climate of the basin. For example, average annual 

 precipitation ranges from less than 15 inches near the headwaters to 

 about 23 inches in the lower part of the valley. The Niobrara drainage 

 basin comprises a narrow strip of land, never in excess of approxi- 

 mately 60 miles in width, across the northern edge of Nebraska with 

 extensions west into Wyoming and north into South Dakota. It may 

 be anticipated that cultural remains will reflect the climatic and eco- 

 logical variations along the east-west profile provided by this basin. 



The 10 reservoir sites reconnoitered, in most instances incom- 

 pletely, in 1950 are all in Nebraska. They are Colwell, in Sheridan 

 and Dawes Counties about 70 miles east of the Wyoming line ; Crooks- 

 ton, Eli, Kilgore, Long Pine, Meadville, Merritt, Sparks, and Thacher 

 in the central section of the basin; and Ponca Creek on the stream 

 of the same name, actually outside the Niobrara drainage but immedi- 

 ately adjacent to its lower reaches. 



Other projects in the basin are the existing Box Butte Reservoir, 

 briefly visited by a River Basin Surveys party in 1946, and two 

 potential reservoirs — Keya Paha and Verdel — not investigated in 

 1950 because their locations were not known at that time. 



Cokvell Reservoir site. — The Colwell Dam, if constructed, will 

 create a reservoir approximately 4^ miles long in the Niobrara River 

 Valley proper and extending some distance up Sand Canyon and 

 Pepper Creek, northern tributaries of the river. The river here is a 

 small stream in a valley of low, flat terraces cut into high rolling plains 

 covered with short grass. Other vegetation consists of scanty trees 

 and brush along the water courses. Part of the reservoir area was 

 examined during a 2-day visit, during which three camp sites and a 

 workshop were found. In each instance, materials were relatively 

 scanty on the surface but this is, at least in some cases, undoubtedly 

 owing to the existence of sterile overburden. The collections made 

 from the sites are preponderantly stone objects, but a sherd or two 

 were found on two of the sites. These pottery fragments are small 

 and nondiagnostic, although one appears to be simple-stamped and 

 accordingly suggests a relatively late date. The three sites which will 

 be destroyed by the reservoir (25SH201, 203, and 204) are believed 

 to warrant test-trenching, and additional survey will be required to 

 complete coverage of the area to be flooded. 



Crookston Reservoir site. — The potential Crookston Reservoir will 

 be a lake approximately I2| miles long if current plans for a dam on 

 the Niobrara River about 10 miles southwest of Valentine are carried 



