328 The American Geologist. NovemiDer, 1902 
changes which preceded and attended the Ice age. While they brought 
greater elevation for northern and western Europe and North America, 
their effect for much of Asia was probably a contemporaneous sub- 
sidence. 
The distribution of the loess in large districts of China seems chiefly 
attributable to wind action, as was first shown by Richthofen and 
Pumpelly ; but in Turkestan, on the northwestern side of the great con- 
tinental plateau, Wright found the wind hypothesis untenable. He says : 
"Evidences accumulated that the land had lately been depressed to such 
an extent that the water of the ocean reached the base of the bordering 
mountains, rising to a bight, certainly, of about 3,000 feet ; for, at this 
level, south and southw-est of Lake Balkash, we found the loess spread 
out in such an extensive terrace that the wind would be entirely in- 
competent to produce the results." 
Marine submergence, thought by the author to have affected a very 
extensive tract of tiorthwestern Asia, would cause the unsubmerged 
eastern part of Siberia to have a moister and warmer climate, giving 
pasturage for herds of the mammoth and woolly rhinoceros far to the 
north, even beyond the Arctic circle. With re-elevation of the land and 
the disappearance of the vast inland sea, the present severe extremes of 
climate would ensue, causing these species to become extinct. 
An excellent chapter on the climate of the Russian empire is con- 
tributed by Mr. F. B- Wright, summarizing the results of weather ob- 
servations during many years at 564 stations, of which 225 are in 
Asiatic Russia, as published two years ago in a most elaborate climato- 
logical atlas by the Nicholas Physical Observatory at St. Petersburg. 
As results of the climate, Siberia consists of three zones, the most north- 
ern being the treeless Arctic tundra ; the next southward is a wide belt 
of forest ; and still farther south are the grassy and flowery steppes, 
similar to our prairies and plains between the Mississippi river and the 
Rocky mountains. w. u. 
Lcs Roches Alcalincs caracterisanf la Province Pctrographiqiic d'Am- 
pasindava, par A. Lacroix. (Nouvelles Archives du Museum 
d'Histoire Naturelle, 4th Series, Vol. I. 4to. 152 pages, ten plates. 
Paris. 1902.) 
Only the first htmdred and fifty pages (premier fascicule) of this 
important contribution to petrography have as yet appeared : but some 
features of interest may be noted briefly — The work is divided into 
three chapters ; the first is devoted to a description of the eruptive 
rocks of Nosy Komba. Madagascar, and their contact phenomena ; the 
second gives the results of study of the rocks of the bay of Ampasin- 
bava and incidentally rocks from a few other localities in Madagascar. 
The third chapter is devoted to the deductions drawn from these 
studies and the theoretical discussions of a general interest to which 
they lead. The last chapter and part of the second have not yet ap- 
peared. 
It follows clearly from the descriptions published that the northwest 
part of Madagascar is a province characterized by a great variety of 
those highly sodic rocks which are generally so rare, and which have 
