352 THE AMERICAN NATURALIST. (VOL: XXXIII" 
r PETROGRAPHY. 
Experimental Petrography. — Morozewicz' has just published a 
long paper on “ Experimental investigations upon the formation of 
minerals in magmas” that will unquestionably take a place among 
the most important contributions to experimental geology that have 
been made within recent years. The author fused known mixtures 
of the various rock-producing compounds in a glass-furnace and 
thoroughly studied the resulting products. The conclusions reached 
by him are full of suggestiveness. The way seems to have been 
opened for a long line of important investigations to follow, some of 
which have already been entered upon. 
The details of the experiments cannot be entered upon here, but 
some of the conclusions arrived at may be briefly indicated. 
1. The structures of cooled magmas appear to depend upon 
exterior conditions of crystallization and upon their chemical com- 
position, quantitative as well as qualitative. 
2. The order of crystallization is determined by no one condition, 
such as fusibility, acidity, etc., but it depends upon a number of 
variable conditions, one of the most important of which is the quan- 
tity of the various constituents present as compared with their 
solubility in the molten mass. The ability of a substance to 
supersaturate the magma depends primarily upon its nature, the 
nature of the other substances composing the magma, and its 
temperature. 
So far as the experiments touch upon the question of the 
differentiation of magmas, they seem to indicate that a molten mass 
may separate into layers or parts differing in density, and that this 
difference may be due to the fact that the bases FeO, MgO, CaO, 
separate as silicates by crystallization earlier than the remaining 
constituents. 
Two pounds of granite, with the composition given below (I), were 
heated for five days. The mixture yielded a mass of glass which in 
its upper portion contained unmelted quartz grains and a consider- 
able quantity of tridymite. Though the glass between the quartz 
grains in the upper portion of the crucible presented the same 
appearance as that in the lower parts, the composition of pieces taken 
from the two parts was found to be quite different. Under (II) we 
1 Min. u. Petrog. Mitth., vol. xviii, pp. 1 and 105. 
