SCROPE — ON INTERNAL STRUCTURE OF GNEISSIC ROCKS. 365 
Mr. Sorby's experiments, in wliich he produced slaty lamination in 
a mixture of pipe-clay and glue containing flattened particles of mica 
or oxyde of iron, by subjecting it to a squeeze, exhibit the same 
result. 
AVe may now proceed to apply these considerations to the case of the 
crystalline rocks of igneous origin. It has been calculated that granite 
must lose one-tenth of its bulk in the process of crystallization or con- 
solidation from a state of fluidity. Then, on the other hand, it must 
augment in volume by one- tenth on passing from a solid to a fluid state. 
Whether tliis be the precise proportion or not, it is quite certain that 
the changes of temperature to which granite and all other igneous rocks 
have been exposed, must have been accompanied by a corresponding 
change of volume, even while its component- minerals still retained 
their crystalline state. But every such dilatation or compression must 
occasion a considerable amount of motion and consequent friction of the 
component crystals inter se. If, as there is great reason to believe, 
water has been always present throughout the crystalline igneous rocks, 
holding much silex in solution, perhaps in a gelatinous state, this will 
have acted as a lubricator to the more solid and still crystalline 
minerals, and enabled them to move among themselves in a consider- 
able degree without being completely broken up, as might be otherwise 
expected, from the amount of friction to which every change of volume 
must have subjected them. Under such circumstances, we can conceive 
a mass of crystalline rock to undergo considerable dilatation through an 
increase of temperature without much change in the form or position of 
its component crystals, on the supposition that the pressures to which 
it is exposed be nearly equal on all sides, and consequently its expan- 
sion nearly, equal in all directions ; although, in this case, many frac- 
tures, cracks, and crevices occasioning a brecciated or veined structure 
may be produced. But, should such an expanding mass force its way 
upwards through a rent broken across a vast weight of overlying strata, 
tilting these latter and shouldering them off on either side, the upper 
and lateral portions of the elevated mass must be subjected to such an 
oblique and unequal strain or squeeze between the downward pressure, 
and perhaps lateral movements, of the tilted strata and the upward 
thrust of the axial crystalline mass, as must tend to break up more or 
less the component solid particles or crystals of these lateral portions, 
and arrange their fragments with their longest axes in the direction of 
