188 RESEARCH IN CHINA. 
deposit so characterized by loess as a chief constituent that it might be 
called Huang-t’u. Von Richthofen, in his hurried journey across the Wu- 
t’ai-shan during a December storm, was led to think that loess was found in 
every corner of these mountains, but he was misled by a few very limited 
occurrences. One of these is at the forks of the T’ai-shan-ho, just above 
Shang-ho-miau, where, in the eddy between wind currents blowing down the 
two valleys to their junction, there is a small deposit on the intervening 
spur. It is necessary to exaggerate the area in order to represent it on the 
geologic map. Another body of Huang-t’u is found just north of the village 
of Wu-t’ai-shan, in the acute angle between two gulches. It is of small 
extent and 25 to 30 feet, 7.5 to 10 meters, thick. The material is very 
fine and free from grit, and a careful search failed to reveal any pebbles 
or even large sand grains. It possesses the vertical cleavage, color, and 
other characteristics of the Huang-t’u of the Great Plain. We regard this 
deposit as a wind drift, owing its accumulation to special topographic 
conditions that induce local drafts and eddies. There are no doubt others 
like it which did not come to our notice. Again, a body of material having 
the characteristics of the Huang-t’u was observed in a hollow on the south- 
east slope of Nan-t’ai, at an elevation of about 6,500 feet, 2,000 meters. 
It may be a remnant of an early deposit antedating the present altitude 
of the region. As is stated in the discussion of the physiographic stages 
(Chapter XI) we are of the opinion that much of the area of the Wu-t’ai- 
shan was formerly covered with Huang-t’u at a time when the altitude of 
the district was very moderate, and it is not surprising that some remnants 
of the formation should remain, possibly in situ in an ancient valley as on 
Nan-t’ai, or as redistributed and redeposited bodies of the indestructible 
material in such situations as that near Shang-ho-miau and near the village 
of Wu-t’ai-shan. 
Descending southward from the Wu-t’ai-shan, we entered the north- 
ern Loess Basins. Near the village of Liu-yiian (atlas sheet C I) we first 
encountered a notable development of the Huang-t’u formation, and thence 
throughout Shan-si it was a dominant feature of the surface. In the basin 
of Téu-ts’un from Liu-yiian to Ts’ai-shi-ling, the formation is spread as a 
gently sloping floor, which reaches upwards on the hills bya gradual curve 
that becomes tangent to their bare slopes. It is scored by innumerable 
canyons from 20 to roo feet, 6 to 30 meters, deep and is clearly revealed 
in section in their nearly vertical walls. In the center of the valley it is 
traversed by the beds of the tributary streams, which bring down great 
quantities of gravel. Sometimes the channel is deeply incised (Fig A, 
Plate XXV); elsewhere the flood waters spread widely over the plain, as 
illustrated in the panoramic view on atlas sheet DI. ‘The distribution of 
Huang-t’u in this valley raises the question of the relative efficiency of 
wind and water in transporting and distributing it. It is clear that on the 
