288 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1926 
at times rain and stream action were important factors in spread- 
ing the loess over the country. The wind-blown deposits would nat- 
urally form during the arid periods that came after the more humid 
glacial times. Osborn therefore places the date of Western loess 
accumulation during the interglacial stages following the second and 
third great advances of the ice-sheet and still more after the final 
retreat of the ice. 
LOESSLIKE FORMATIONS IN CHINA 
A. OLDER THAN THE LOESS 
1. Hipparion beds—Previous to the careful work of Andersson 
and Zdansky, the upper and lower limits of the loess had not been 
recognized. Hence the distinction between it and the more 
reddish clay underlying it in many localities passed unnoticed by 
v. Richthofen and by many others since. In fact, the great explorer 
notes, as typical of the loess, the presence of large lime concretions 
which in point of fact are much more characteristic of the red clay. 
Sowerby (2-116) in the report of the Clark Expedition of 1908-9 
recognizes the red clay as a distinct unit (shao ?w), but having no 
fossil evidence on which to separate it, classed it, as did Willis, 
with the loess. Perhaps this in part accounts for the extreme thick- 
ness he attributes to the loess.$ 
On the north margin of the loess basin the name is perhaps mis- 
leading, as here the “red clay ” is not a true clay and is often only 
reddish brown in comparison with the yellow gray of loess. In 
the center of the area it is more true to its name. It has the appear- 
ance of a residual soil resulting from the decomposition of rocks 
an situ. Moreover, Andersson has noted that its distribution is 
practically limited to the old limestone lands. This has led him 
to suggest that though the bulk of the loess came as dust blown from 
the desert in the manner suggested by v. Richthofen, much of it may 
have been material formed locally and only re-sorted or slightly 
shifted by the wind. 
Locally the red clay shows a poorly developed stratification marked 
by gravel beds, may reach a thickness of 200 feet, and in several lo- 
calities, which are thought to have been oases, has yielded rich 
collections of animal remains. According to Drs. Wiman and 
Zdansky the animal types of this Hipparion fauna® indicate the 
existence of steppe conditions in China at the close of Miocene and 
the beginning of Phocene times. 
8The loess covers the whole of the sedimentary rock to an average depth of 1,000 
feet * * * south of the Ordos Desert the depth increases to 2,000 feet (2-128). 
® Andersson (1-109) gives a provisional list of the animals found by Zdansky in one 
locality, which includes, among others, deer, antelope, boar, fox, hyena, saber-toothed 
tiger, elephant, mastodon, and turtle, together with Hipparion richthofeni, Aceratheriwm, 
Stegodon, and the ostrichlike bird referred to already. 
