96 
IDDINGS. 
a molten rock magma is nothing more than a supersaturated 
solution of different silicates, which only need a slight in¬ 
centive, according to the degree of saturation of each com¬ 
pound, in order to crystallize out as rock constituents. He 
discusses at length the crystallization of magmas from a 
chemical standpoint, and the influence of his views upon 
the present paper may be recognized in many places. 
In the light of our present knowledge, or in the absence 
of any knowledge to the contrary, this seems the most reason¬ 
able hypothesis as to the nature of rock magmas. 
It is to be remarked that in considering these magmas 
with respect to their crystallization we have only to deal 
with them near their point of solidification, and therefore 
have no need of considering their possible condition at ex¬ 
cessive temperatures. 
It is also to be remembered that the actual chemical status 
of solutions and the physical conceptions of fusion and solidi¬ 
fication are as yet unsolved problems of the most essential 
nature. Nevertheless, some of the laws governing the crys¬ 
tallization of solutions have been investigated to a certain 
extent, and have been formulated by Sorby.* He found that 
the crystallization of those salts that condense upon crystal¬ 
lization from a fluid condition is aided by an increase of 
pressure, and that the reverse is true for those which expand 
upon crystallization. The silicate rock-making minerals are 
found to belong to the former class, for their crystallized con¬ 
dition is denser than their amorphous or glassy modification, 
and these glasses are denser than the molten substance at the 
same temperature. Hence it is assumed that when they exist 
in a fluid magma a diminution of pressure would lower the 
point of temperature at which they would crystallize out of 
the magma; or, expressed differently, if a magma were on 
the point of crystallizing under a particular pressure and 
temperature, a diminution of the pressure would prevent the 
crystallization by decreasing the degree of saturation, and 
*Proc. Boy. Soc., London, vol. 12, 1863, p. 538. 
