50 PROOEEDTNGS OF THE r'llICAGO MEETING 



iiR-lu(U'(l ill Carman's teri-itory or in the area orijiinally oxiimined liy the 

 writer. The gravels in this original area were found in valleys cut in Kansan 

 drift, and so belong in the same series as Carman's Valley gravels, which he 

 defined as being "those gravels which occupy valleys in the Kansan area that 

 could not have been reached by outwash from the Wisconsin ice-sheet, and 

 which are, therefore, not Wisconsin gravels." This definition has been adhered 

 to in the present study, although it may be added that, even in the case of 

 valleys which could have been and probably were reached by Wisconsin flood- 

 waters, the true valley gravels, according to the above definition, are distin- 

 guishable by being overlain by loess. 



It is a characteristic of northwest Iowa that the Kansan till is fresh nearly 

 or quite to its surface, in contrast with the Kansan till of southern Iowa, 

 which Is leached for several feet below its upper surface. This characteristic 

 seems to imply that the upper leached portion of the till was eroded prior to 

 the deposition of the loess. It is probable that the valley gravels owe their 

 origin to this erosive process, which was concomitant with a differential uplift 

 of this part of the State. 



Every stream valley heading within this area of fresh till, so far as these 

 have been examined, contains valley gravels. These are in great part over- 

 lain by loess or a loess-like clay. The same is true also of valleys in west- 

 central Iowa which head south of the area which was so distinctly affected 

 by sheet erosion as to have all its leached till removed. The west margin of 

 the Wisconsin drift-sheet is close to the Mississippi-Missouri divide in this 

 region, so that few streams in the Kansan area drain eastward. However, 

 one such valley was examined and was found to contain abundant -valley 

 gravels. Nearly all the streams of northwestern Iowa flow southeast or south- 

 west, but at least one northward flowing stream is bordered by valley gravels 

 as abundantly as are those streams flowing in the normal directions. 



As to the date of their formation, these gravels are, of course, pre-Peorian, 

 since they are overlain by loess which is recognized at present as being of 

 Peorian age. How much older they may be is uncertain. As they are un- 

 leached and unweathered and are in places interleaved with loess, they would 

 seem to be but little older, though this can not be positively afiirmed. They 

 are doubtless derived chiefly from the Kansan till. They must be differen- 

 tiated from the so-called Aftonian gravels, which in the type localities Kay 

 has shown to be really included masses of gravel rather than interglacial 

 horizon markers. They are also distinct from the gravel hills and masses of 

 northwestern Iowa, which Carman showed were likewise included within the 

 Kansan till. These included masses must have contributed largely to. the 

 valley gravels. The valley gravels lie indiscriminately on Kansan or Nebras- 

 kan till and are found at different heights above the streams, ranging from 

 water level to fully eighty feet. 



Read from manuscript. 



Discussed by Prof. J. A. Udden, George F. Kay, and J. Ernest Carman. 

 Avith replies by the author. 



