53§ THE JOURNAL OF GEOLOGY. 



the glaciated region of North America is believed to be the 

 product of glacial waters, it still remains, in my view, probable 

 that certain parts of it were deposited by winds. This part is 

 believed, in general, to have been derived from the water- 

 deposited portion, but perhaps this is not universally true. 

 Along the leeward side of the Mississippi river, for instance, we 

 find dunes of sand and dune-like accumulations of loess that 

 seem in both instances to have been derived by winds from the 

 flooded flats of the river below. In like manner, there seems 

 ground for the belief that in Pleistocene times, the glacial floods 

 alternately extended and withdrew themselves, leaving great silt- 

 covered flats exposed to wind action, and that from these silt 

 was swept up and deposited over adjacent and perhaps somewhat 

 distant highlands. It seems also not improbable that the con- 

 ditions of the surface may have been such as to permit the lodg- 

 ment of this more uniformly over the surface than is the habit 

 with dunes. There seems ground for this in the distinction 

 between the formation of dunes and the supposed deposits. 

 Dunes are formed from sand driven along the surface by winds, 

 but not in any notable degree carried by the winds in full sus- 

 pension. The supposed silt deposits, on the other hand, are 

 presumed to have been formed by silt borne in free suspension 

 until, by contact with the earth, it was lodged. Such contact 

 might obviously be widespread and the lodgment product might 

 have a wide and measureably uniform distribution. While, there- 

 fore, coinciding with what seems to .be the majority opinion 

 among American geologists that the loess deposits of the 

 glaciated region are chiefly water-lain, it appears to me prudent, 

 if not important, to recognize the asolian class, and to search 

 diligently for criteria of discrimination between the two classes. 

 The foregoing classification is consciously incomplete. In 

 some instances the bases of distinction border closely upon the 

 structural rather than the genetic, but it is believed that there is 

 involved in every case an important genetic factor, though it 

 may sometimes be most conveniently expressed in structural 

 terms. T. C. Chamberlin. 



