170 The American Geologist. March, is'JT 
As the magma cools, the various minerals develop and crys- 
tallize out in the general order of their basicit}', each, in turn, 
excluding, so far as possible, the still unindividualized magma 
and water, just as any crystals formed in a solution exclude 
the solvent. The magma thus becomes steadily more and 
more acid and hydra ted; and when, finally, the free silica 
alone remains unsolidified, it is probably combined with suf- 
ficient water to hold it temporarily in the gelatinous state. 
As this gelatinous silica slowly and spontaneously dehydrates 
and crystallizes it necessarily imprisons the main part at least 
of the water, which, like a true solvent, is the only part of 
the magma that remains permanently unsolidified. 
Conceive, now, that a large boss or massif of molten gran- 
ite has invaded the crystalline schists at a great depth and is 
undergoing extremely slow refrigeration. Crystallization be- 
gins on the surface of the mass and gradually extends inward.. 
The rate of cooling diminishes as the crust thickens, but the 
highly viscous nature of the cooling magma prevents the for- 
mation of large crystals. Finally, however, as the outer shell 
of the boss becomes gradually thicker, the rate of cooling and 
crystallization are sufficiently reduced to enable the quartz, 
as it slowly hardens, to exclude a portion of its inclosed water, 
forcing it inward into the still liquid magma; which is thus 
rendered more liquid, but not sensibly so at first, for the slow- 
ness of the process permits the ditfusion of the excluded 
water through a large volume of unconsolidated magma. The 
extremely gradual nature of the changes is an essential fea- 
ture of the explanation. The feldspar and mica probably 
crystallize in part at least, far in advance of the quartz, 
forming a reticulated or sponge-like zone, through the meshes 
of which the silica gradually expels a portion of its water. 
This process is, perhaps, favored by the tendency, through 
contraction, to reduce the pressure in the interior of the mass. 
The magma thus becomes, although its temperature is slowly 
falling, more and more liquid, passing graduallj^ from the con- 
dition of aqueo-igneous fusion to that of igneo-aqueous solu- 
tion ; and the greater mobility of the latter state permits the 
formation of larger crystals. We may thus conceive a perfect 
blending of the conditions favoring the formation of normal 
granite in the exterior portion of the mass with those favor- 
