QUATERNARY. 193 
wind sorts and redistributes when the waters are low. This alternation of 
activities occurs during the rainy summer season and the dry winter season, 
in a manner which makes the winds especially effective, since at the time 
when they are active the ground is denuded of vegetation and unprotected 
by snow. 
The level plain of the Hin-chdéu basin affords no sections of the under- 
lying Huang-t’u formation, and it is not possible to determine whether 
the material be stratified or not. We can only infer from the adjoining 
basin of Huang-t’u-chai on the south that the formation may be stratified 
as itis there. Along the base of the Ki-chéu-shan the Huang-t’u formation 
takes on a peculiar character, incident to the close proximity of the great 
fault-scarp. Along a portion of the base of the range alluvial cones rise to 
a height of 1,000 feet, 300 meters, above the plain and have a radius of 
from 2 to 3 miles, 3 to 5 kilometers, from the mouths of theravines. With 
this great body of mountain wash is mingled the loess that is blown in 
quantities from the plain onto these lower slopes. We thus get a com- 
posite formation, in which sometimes the gravel and sometimes the loess 
predominates. 
The Huang-t’u formation covers the Shi-ling pass, extending in a 
continuous sheet from the Hin-chéu basin to that of Huang-t’u-chai. The 
slope rising from the Hin-chéu basin to the Shi-ling resembles the curve 
by which the surface of the plain elsewhere passes into the declivities of 
the hills, and we recognize the form which is given by the drifting and 
scouring action of the wind. The slope is also scored by ravines, which 
display the vertical structure of the Huang-t’u, but if it was stratified we 
did not note the fact. On both sides of the pass itself the road lies in a 
cut from 20 to 60 feet, 6 to 18 meters, deep, which also exhibits a body of 
homogeneous loess. A single limestone outcrop protrudes through the 
general mantle near the center of the pass, and hills partly clothed with the 
Huang-t’u rise both east and west. There is no positive evidence of the 
fluviatile origin of the deposit in the pass, neither can we readily attribute 
it to the wind, since the situation is one which is subject to scour and 
unfavorable to accumulation. On the ground that the unfavorable con- 
ditions practically preclude eolian deposition, we incline to think that the 
body of material in the pass is continuous with that in the Hin-chéu basin 
and with that in the basin of Huang-t’u-chai, and is of alluvial origin. Its 
present altitude may be interpreted as an effect of the upwarp of the 
Ki-chéu-shan, to which the diversion of the Hu-t’o-ho from its former 
course through this pass is attributed. The surface forms of the slope, 
rising from the Hin-chéu basin and extending across the pass, are effects 
