IOWA ACADEMY OF SCIENCES. 47 



bulk, but was a gradual accumulation such as would 

 appear impossible in water in view of the extent and pres- 

 ent topographic features of the region in which the deposit 

 is found. 



The influence of plants upon the formation of the loess 

 is also suggested by the fact that the deposit is thickest on 

 highlands, either adjacent to streams or along morainic 

 lines. On these highlands terrestrial vegetation was first 

 able to establish itself after the recession of the glaciers 

 and while the lowlands were still covered by water at least 

 a part of the year, and in such places taller vegetation has 

 been able to maintain itself to this day, especially in close 

 proximity to streams. 



Such highlands would also lie near the chief source of 

 supply of the fine materials, already discussed by the 

 writer. * 



Since the retreat of the ice-sheet our rivers have been 

 subject to overflow, especially during the spring, after the 

 melting of snow and ice. During the flood period large 

 quantities of detritus have always been transported, and 

 finally deposited in bars of mud and sand. During the 

 summers the waters have subsided and exposed these bars 

 to the summer winds. Most of the material taken up by the 

 winds was deposited among the plants on the skirting hills, 

 where it entombed the shells of terrestrial snails; some, 

 especially the finer particles, was carried farther and 

 deposited in part among the more lowly plants of the 

 prairie, among which snails are seldom found. The 

 morainic highlands were situated in much the same posi- 

 tion. They were adjacent to streams in most cases, and 

 sometimes to ponds and lakes which furnished very similar 

 conditions. 



Because loess was abundant along the borders of the 

 lowan drift it was concluded that glacial agencies played 

 an important part in furnishing material for the deposit. 

 The significance of this circumstance has probably been 

 misinterpreted. If the border of the lowan drift be traced 

 from the Wisconsin margin in Hardin county to Scott 



*See ' -The Loess of Natchez, Miss. "-Am. Geol. , Vol. XXX, pp. 295-6, 1902. 



