238 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. Q3 



chronology of postglacial changes in climate and vegetation in eastern 

 North America (Sears, 1932, p. 621) is not possible. However, 

 in a later letter "" he states that if the material immediately below 

 level I is old loess, then that horizon could be correlated with the 

 cool, dry conditions previous to about 7,000 years ago. (Sears, 1932.) 

 On the other hand, if the deposit (in and) below level I is definitely 

 not aeolian, then (a) either the period before 7,000 years ago was not 

 favorable to loess formation, which he doubts, or else (b) level I is very 

 old, having been formed at or j^revious to 9,000 years ago, when 

 conditions were cool and moist. (Compare Sears, table 3, 1932. p. 

 621.) In the latter case, level 11 is the only trace of the humid in- 

 terval separating the cool dry from the warm dry periods, and is 

 about 5,000 years old, with level I almost twice as old. From a photo- 

 graph, he is inclined to believe that there is old loess below and in 

 level I. Dr. W. Van Royen, who has visited the butte, but has not 

 seen the best exposures of this sand and gravel layer (fig. 29) under 

 level I, is likewise unconvinced of the fluviatile origin of this deposit. 

 (Personal conversation, January 1934.) On the other hand, Kirby's 

 detailed study and mechanical analyses of this material have convinced 

 us of the aeolian origin of level I (aside from human detritus) and 

 of the water-borne nature of the silt and sand immediately beneath 

 it. These sorted sand beds and silts occurred under level I through- 

 out our excavations, especially in the west-central portion of the butte, 

 and a test pit in the cast end showed especially thick graded quartz 

 sands in that area. The question can only be settled by full publica- 

 tion of the data and through checking by more competent authorities 

 than Kirby and myself. This sand stratum, being barren of artifacts, 

 has probably not been ruined by relic hunters. Moreover, there are 

 at hand definite soil samples from all these levels. This is not directly 

 an archeological problem but is presented here because of its probable 

 bearing on dating the horizon and as an example of the manner in 

 which prehistoric archeological chronology must always depend on 

 the researches of geographer and geologist. 



We may say then, subject to later correction, that level I on Signal 

 Butte probably falls either in the dry, cool i)eriod (circa 8,000 years 

 ago) or in the humid cold period (circa 10,000 years ago). In 

 general, Kirby's time estimate of dune migration agrees. His esti- 

 mate of 5,200 years as a probable minimum must be ex])anded if 

 humid periods produced slackening rates, and there is evidence that 

 humid periods occurred during the formation of the earth cap. This 

 factor permits the general correlation of his minimum estimate with 



Paul B. Sears to Maurice I'". Kirby, March 25, 1933. 



