S6 The Amerkan Geologht. Juiy,i893 
from it as are the rocks of New England, which liavo been swept 
clear of all such material by the recent ice advances. More- 
over the bare granite rock contains depressions seemingly due to 
erosion, without outlet, and varying in size from a few yards to 
several thousand feet,* and apparently' the result of ditferences 
in rock texture. Such depressions would be expected to exist at 
the base of an area of secularly decayed material. From these 
facts the author argues that we may have here a soui*ce for the 
assumed aeolian deposits of the loess. If it is assumed that 
China was exposed to the conditions which southern Asia has 
undergone, the country must at one time have been covered l)y 
■a well marked residual soil. None now remains and it must have 
been swept away by some agent. Not by the sea, for there is no 
evidence of a marine incursion, nor by ice action, for evidences 
•of the existence of glaciers are entirely wanting. The agency of 
the wind alone appears to remain to account for this change. A 
change in climatic conditions from moist to diy would set to 
work this powerful agent of transportation. 
By such a transformation a series of deposits are rendered pos- 
sible. The finest material driven great distances by the wind 
would finally settle and remain and accumulate under the protec- 
tion of the steppe grasses in the region intermediate between the 
arid interior and the moist coast climates, forming a deposit not 
unlike loess; the coarser sands wovdd form the drifting desert 
sands; and the still coarser parts would remain liehind and form 
the characteristic coarse stony steppes. The conditions in China 
seem to suggest the plausibility of this explanation. 
Pumpelly offers a similar explanation for the loess of the trans- 
Mississippi regions. During early times the region of the Rocky 
mountains must have been the seat of secular disintegration, though 
now none of its products remain in place. The present aridity and 
prevailing west winds may account for its removal and for the 
presence of loess as far east as Kansas. The loess deposits of the 
Rhine and of the Central states may have been accumulated by 
feolian action, but in these cases the supply may more likely have 
been glacial flour which itself may have lieen derived by the gla- 
cier from secularly decayed soils. 
*Pumpelly, Geological Researches in China, etc., Smith'n Contr. 
Knowledge, xv, Art. 4, 2S, 72, 73. 
