1 882.] The Loess of North America. 379 



forms in the loess regions on the borders of Mongolia. The 

 points, however, which are mainly relied on by the Baron, appear 

 to be (1) the presence of root-marks occurring throughout the 

 formation ; (2) absence of fresh-water or other aquatic life-forms ; 

 and (3) absence of stratification. Both the second and third of 

 these propositions are met by the repeated statements of numer- 

 ous careful observers, who have found aquatic and semi-aquatic 

 forms in many localities. The presence of the semi-aquatic 

 forms alone — such as Succinea — which are indicative of a moist 

 station, effectually negatives the assumption of a " dry, elevated 

 area swept by fierce winds." The first proposition has been met 

 by the studies of Professor J. E. Todd, 1 who has shown that from 

 the law which evidently obtains, that root-marks vary in fre- 

 quency inversely as their distance below the present surface, " un- 

 usual care is necessary to interpret observations correctly." The 

 conclusion reached by that observer is, that when correctly inter- 

 preted the distribution of root-marks opposes the sub-aerial hy- 

 pothesis. • 



It will have been observed that the original statements and in- 

 ferences of Pumpelly and those of Richthofen were distinct and 

 opposed. The former recognized the agency of water alone as 

 sufficient to explain the phenomena he studied, while the latter 

 called to the aid of the winds a lake-basin, which in turn neces- 

 sitated his artificial distinctions of lake-loess and land-loess. 

 That such a distinction is wholly inapplicable to American de- 

 posits — unquestionably true loess — is patent, for the reasons that 

 it presents a perfect homogeneity of structure, entire absence of 



structures — and does present at several localities a faintly strati- 

 fied appearance. Besides, the climatic conditions required by this 

 theory of the Chinese loess, seem to have had no counterpart in 

 climatic changes over the areas covered by our loess. 



The argument for the lacustrine origin of the American deposits 

 has been in part anticipated in the foregoing. But there should 

 be added the facts that here the formation is confined to river 



