NO. 10 NEBRASKA ARCHEOLOGY STRONG 177 



loess such as could occur only as wash from the hills. The distribution of these 

 various materials in the gully and for some distance in its banks (as determined 

 by boring with a long auger) makes it clear that the present stream is cutting 

 across the course of an older stream whose channel was wider than that of 

 the present creek. The banks of the older stream valley had a much gentler 

 slope than those of the present gully, and its course was not the same as that 

 of the present stream. Where the latter cuts the old channel the gully walls 

 consist of " gumbo " clay, gravel, or secondary loess. Where it does not cut 

 the old valley the walls are of the original yellow clay. 



The general situation above described is shown on a rough sketch 

 map made by the present author, February 14, 193 1 (fig. 24). 



On the flats around the gully Sterns found numerous potsherds and 

 a few flint implements. We will describe this pottery later, but for 

 the present it will suffice to say that in the main it resembles the 

 ceramic remains from the rectangular earth lodges (i. e., the Nebraska 

 culture). As the flat has been plowed since 1857, at present no visible 

 traces of house pits remain, but there was a remarkable concentration 

 of pottery on the flat compared to its scarcity in the gully of Sterns 

 Creek itself and its total absence on the surrounding ridges. These 

 observations made by Sterns were corroborated by our own work 

 here in 1930 and 1931. Hence he argues with reason that the rec- 

 tangular-earth-lodge people dwelt on the flat itself and not on the 

 ridges above. Since there is thus no evidence that the present surface 

 cultural debris was washed down from the higher ground we can 

 safely assume that the present surface of the flat on both sides of 

 'Sterns Creek represents the living level of the prehistoric rectangular- 

 house people. 



Coming to the gully of Sterns Creek itself, we find it to be generally 

 straight-walled and, according to Sterns in 191 5, averaging 20 feet 

 in depth. It may be significant that our own measurements in 193 1 

 showed a depth of 25 feet 6 inches at exposure 5 (map, fig. 24) and 

 of 24 feet at exposure 8. Whether this actually represents an addi- 

 tional 4 feet 6 inches of cutting since 191 5 can be determined only 

 by detailed comparison of Sterns' data (1915 a, pp. 179-187) with a 

 complete series of measurements taken at the points indicated in his 

 description. However this may be, we may note that the cutting of 

 Sterns Creek has exposed human habitation levels in the lower por- 

 tions of the gully walls. Sterns notes that 4 feet below the present 

 surface there are traces of charcoal but no artifacts. He points out 

 that this may indicate casual human habitation at this level, but, 

 judging from the very few scattered pieces of charcoal personally 

 observed in the banks at about this depth, I would hesitate to go so 

 far. Certainly there are no fireplaces and even very little charcoal ; 



