OF VOLCANIC ROCKS. 53 
We may then distinguish as eminently probable, the following immediate effects 
of the cooling of the globe: 1. Contraction of the liquid portion of the interior, by 
cooling down to the temperature required for solidification under the respective press- 
ure. 2. Expansion by slow and perfect crystallization. 3. Contraction of the erys- 
tallized masses by further cooling. We may add to these as a secondary effect of 
greatly inferior importance, the flow of heat or change of chthonisothermal planes, at 
those places where subsidence causes an accumulation, or elevation an abrasion of 
matter on the surface. This process, to which a truly wonderful importance has been 
ascribed by some, must be going on continually and everywhere. Being itself the 
result of motions of the crust, its proximate cause must be the processes attending the 
cooling of the globe, while in its totality, it may have a further, though very insignifi- 
cant, effect in shifting the areas of elevation and subsidence. 
Other agencies, with which we are yet unacquainted, may probably result from 
the same common cause. But taking into consideration only those already mentioned, 
it may appear difficult to form a clear conception of the mode in which they must 
codperate, in order not to counteract each other, and thereby to result in a general 
movement of the crust in one direction only, but to bring about that variety of 
effects which manifests itself chiefly in the change of elevation and subsidence, and in 
other modes of dislocation of the crust. It is perfectly evident, that the aggregate 
effect of those agencies is different under different portions of the crust of the globe, 
contraction prevailing in some parts, and expansion in others; the former causing 
subsidence and deposition of sedimentary matter, the other elevation and denudation. 
But if we consider how gradual is the passage in the state of aggregation from the liquid 
interior of a current of lava to its solid crust, and if we bear in mind how immensely 
vaster in volume are those masses which constitute the interior of the globe, and what 
immensely longer time is given for their cooling and crystallization, the conclusion is 
irresistible, that the passage in the states of aggregation from solid to liquid must 
extend there over an immensely greater space. Crystallization will very probably 
take place next to those portions of the crust which are already solid. If it is attended 
by an increase of volume, and this increase produces tension, it is very likely that the 
next adjoining masses, though not yet crystallized, will offer too much resistance to 
allow this tendency for expansion to find immediate relief by yielding to the gen- 
eral tendency to contraction which may prevail in liquid masses at greater depths. 
It may thus be explained why, in different parts of the crust, a motion, independent 
and in opposite directions, may result from the two-fold tension attending changes 
of volume which take place at different distances beneath the surface, and which 
and others with a view of ascertaining the increase in volume which different rocks undergo by melting, do not affect our 
supposition. In the first place, the molten state may be quite different from that in which the same substance would be 
immediately before crystallizing, and the passage from one to the other may be attended by a decrease in volume which is 
not counterbalanced by the subsequent increase taking place by the passage of the substance into the crystallized state. 
Then, the rocks on which the experiments were made had been solidified under a great pressure, while, when in a molten 
state, they were only exposed to the pressure of the atmosphere. Finally, it has not been ascertained whether all the water 
which is contained in the rocks does escape on melting or not. If not, then it will probably have the effect of inflating the 
molten mass. 
G (91) 
