LOESS 567 



River, as well as the presence of the loess where the prevailing 

 winds are least likely to pile it up. (2) The exclusive dependence 

 upon wind for the distribution of the loess leaves out of considera- 

 tion the difficulty of accounting on this basis for the extensive level- 

 topped terraces which frequently occur. (3) The agency of water 

 in the distribution of loess is favored by, or at any rate consistent 

 with, the fact that the loess is quite uniformly found to be thickest on 

 the margin of the streams flowing out of the glaciated region, and 

 that it is of somewhat coarser texture nearer the streams, thinning 

 out at a distance from them, and merging gradually into a more 

 clayey deposit. The stratification of the loess is also considered as 

 indicating deposition by water. Boulders and granitic pebbles have 

 been found in the loess on the Osage River, which can be accovuited 

 for only on the supposition that they were carried there by floating 

 ice. The source of the material of the Mississippi Valley loess 

 seems to be in close connection with the deposits of the last glacial 

 period. This is shown not only by the distribution, but also by the 

 not infrequent interstratification of loess and marginal glacial drift 

 and till. The close relationship of the Asiatic and European loess 

 to the ice border has also been suggested, but for Asia this is at least 

 somewhat doubtful. Indeed, the source of a part even of the Amer- 

 ican loess is traceable to the nonglaciated arid regions of the west. 

 Our present knowledge of the character of the deposits desig- 

 nated loess leads to the conclusion that the source of tlie material 

 is various, being in part the dust blown from the deposits in. front 

 of the ice sheet, in part the dust derived from deserts and steppes, 

 and in part the silt of river flood plains or other sources. The 

 glacial origin fits best the American and European, and the desert 

 origin the Asiatic deposits, but none are probably of simple char- 

 acter throughout. Though a marine origin has been advocated for 

 the loess by some writers, the consensus of opinion is that the loess 

 is a continental deposit. The evidence, further, points to the wind 

 as the chief agent in distribution, distribution by rivers being of 

 secondary importance. Deposition was for the most part upon dry 

 land, vegetation playing no doubt an important part in the separa 

 tion of the dust from the air, though rain also must be credited 

 to a certain extent with this function. It is, however, essentially 

 a steppe deposit. All such loess is practically without stratification. 

 .Some loessic material was no doubt deposited in standing water 

 and more perhaps on river flood plains, but the totality of this was 

 far below that of purely eolian origin. Secondary reworking and 

 redistribution l)y wind, rivers and waves also occurred, and is going 

 on at the present time. To a certain extent this is also true of the 



