338 NATURAL HISTORY BULLETIN. 



between the Iowan and the loess of the Missouri and Big Sioux 

 river regions. Indeed, Bain refers to the northwestern loess 

 in the following words:* 



"It is known, however, that loess in northwestern Iowa prob- 

 ably belongs to more than one geological epoch, and Professor 

 Macbride's observations in Humboldt county make it conclu- 

 sive that the Iowan did not cover the region immediately north 

 of Carroll county, as has heretofore been believed. The cor- 

 relation of this loess in Carroll county with the Iowan drift is 

 accordingly open to considerable doubt." 



If these words apply to loess comparatively near to the 

 Iowan border, what shall be said of the great mass of loess 

 along the Missouri and lower Mississippi rivers, covering an 

 area vastly greater than the known Iowan border region, and 

 between which and the Iowan no connection whatever has 

 yet been established ? f 



It is the writer's opinion that the accumulation of a com- 

 paratively large amount of loess along the border of the Iowan 

 drift is explained by the fact that this border follows the 

 larger streams of this part of the state, the Iowa, Cedar, Wap- 

 sipinicon, Maquoketa and Turkey. The deposit is thickest 

 in the southern portions of the area, where the river valleys 

 are broad, and where winds could easily gather up quantities 

 of dust from the sand and mud bars exposed at low water. 

 In any case no proof has yet been furnished that the Lansing 

 loess is contemporaneous with the Iowan drift. 



The attempt which Upham makes]: to divide the loess into 

 an aeolian "upland loess" and an aqueous "valley loess," 

 which is practically a repetition of Hershey's effort, § is not 

 successful. He says: ". . the winds . . . blew away much of 



*Ia. Geol. Sur.,vol. ix, p. 92, 1899. 



tvSee paper on Loess and the Iowan Drift, which follows. 



J Am. Geol, 1. c. , p. 29. 



§See Am. Geol., vol. xxv, pp. 369-374, 1900. 



