100 
IDDINGS. 
magma cools under a variety of physical conditions which 
obtain for different portions of its mass the resulting crystal¬ 
lization should vary with the changes of physical conditions. 
If the variations in the physical conditions which attend 
the crystallization of the great majority of magmas are con¬ 
fined to a certain range the corresponding differences in the 
crystallization will also be limited, and will be recognized as 
of frequent occurrence. Greater variations in the physical 
conditions would lead to less frequent variations in the crys¬ 
tallization, both of which would be considered exceptional. 
By crystallization is to be understood both the kind of min¬ 
erals developed and the structure they assume. 
(3.) These theoretical conclusions agree with the general 
observations already stated, that among rocks of similar 
geological occurrence—that is, where the physical history of 
the crystallization has been similar—the mineral composition 
of the different kinds of rocks varies according to the chem¬ 
ical composition; that among rocks of different geological 
occurrence—that is, where the physical history has been dif¬ 
ferent—the mineral composition of rocks of similar chemical 
character varies according to the geological occurrence. 
To these rules there are many apparent exceptions; but it 
must be remembered that only a portion of the physical his¬ 
tory of the crystallization of any rock can be discovered by 
observation, and that the greater part of it must be derived 
by indirect methods, if at all. Thus it is possible to judge 
in many cases of the conditions under which crystallization 
took place after a magma came to rest from the geological 
environment in which we find it afterwards; but there is no 
direct means of observing what conditions it passed through 
before coming to rest. 
■ 
Consideration of Magmas during Their Eruption .—The gen¬ 
eral nature of the physical changes experienced by a magma 
previous to its coming to rest can be understood if we con¬ 
sider the act of eruption of igneous magmas. 
In this consideration we shall start with a molten, fluid 
