252 RESEARCH IN CHINA. 
covered it, and there seems no doubt but that the huge rocks rode out on 
to the plain on the loess, which could only move in that manner if satu- 
rated, but in that state could flow and transport them, even 2 miles. 
Where the Huang-t’u extends below the level of ground water, as in river 
flood-plains, it no doubt is a reservoir holding a very great volume of 
water in its pore spaces; but when it accumulates above ground water, 
or is drained as that level is lowered by lowering of the point of discharge, 
then it usually remains unsaturated, though perhaps moist from top to 
bottom. 
The vertical cleavage of the Huang-t’u is one of its most general 
and conspicuous characters. The view expressed by von Richthofen and 
accepted by Pumpelly, that it is due to pores left by the decay of grass 
roots, may hold in districts where the steppe grasses prevail and for deposits 
on which they have grown continuously, but it does not suffice to explain 
the structure in masses laid down under the varied conditions of the Huang- 
t’u in China: where climatic conditions have been from time to time more 
or less arid; where the surface has at times been that of a drift or of a 
flood-plain, or of a lake bottom; and where the structure extends from 
top to bottom of every mass, however thick, and is coextensive with 
the formation over hundreds of square miles. Grasses which should be 
so universal and so persistent as is this structure would excite wonder. 
We saw the vertical cleavage in masses that were stratified, as well 
as in those that were not stratified; in water-laid and wind-drifted deposits; 
in plains and on mountain slopes; in older and younger accumulations. 
It thus appears to be independent of conditions of deposition, site, and 
age. It does not occur in unconsolidated drifts of recent dust storms; 
it is uncertain whether it would be found in Huang-t’u which extends 
below ground water level, but it does occur in dried, 7. e., consolidated, 
deposits of Huang-t’u, of whatever genesis. It is better developed in wind 
drifts which are composed wholly of loess than in water-laid bodies that 
include appreciable amounts of unsorted residual clays; thus it seems to 
be related to texture. Texture and consolidation seem to be the governing 
factors in its development. Von Richthofen says:* 
There can be no doubt that this remarkable phenomenon of the tendency of loess to 
a vertical parting has its origin in the peculiar vertical capillary texture. 
Agreeing with him in this conclusion, but finding the theory of grass 
roots altogether inadequate, I suggest the following activity of physical 
and chemical processes induced by gravity, capillary attraction, and cemen- 
tation, through the medium of moisture and the fine earthy particles. A 
* China, vol. 1, p. 61. “Es kann keinem Zweifel unterliegen, dass dieses merkwiirdige Phanomen, 
der Tendenz des Léss zu einer verticalen Absonderung, in der eigenthiimlichen verticalen Capillartextur 
seinen Ursprung hat.” 
