a mass of strata,
when heated, and therefore in some slight degree at least softened,
would bow outwards like a bar of metal. Consider of how many subordinate
layers each great mass would be composed, and the mineralogical changes
in any length of any one stratum: I should have thought that the strata
would in every case have crumpled up, and we know how commonly in
metamorphic strata, which have undergone heat, the subordinate layers
are wavy and sinuous, which has always been attributed to their
expansion whilst heated.

Before rocks are dried and quarried, manifold facts show how extremely
flexible they are even when not at all heated. Without the bowing out
and subsequent filling in of the roof of the cavity, if I understand
you, there would be no subsidence. Of course the crumpling up of the
strata would thicken them, and I see with you that this might compress
the underlying fluidified rock, which in its turn might escape by
a volcano or raise a weaker part of the earth's crust; but I am too
ignorant to have any opinion whether force would be easily propagated
through a viscid mass like molten rock; or whether such viscid mass
would not act in some degree like sand and refuse to transmit pressure,
as in the old experiment of trying to burst a piece of paper tied
over the end of a tube with a stick, an inch or two of sand being
only interposed. I have always myself felt the greatest difficulty in
believing in waves of heat coming first to this and then to that quarter
of the world: I suspect that heat plays quite a subordinate part in the
upward and downward movements of the earth's crust; though of course
it must swell the strata where first affected. I can understand Sir
J. Herschel's manner of bringing heat to unheated strata--namely, by
covering them up by a mile or so of new strata, and then the heat would
travel into the lower ones. But who can tell what effect this mile or
two of new sedimentary strata would have from mere gravity on the level
of the supporting surface? Of course such considerations do not render
less true that the expansion of the strata by heat would have some
effect on the level of the surface; but they show us how awfully
complicated the phenomenon is. All young geologists have a great turn
for speculation; I have burned my fingers pretty sharply in that way,
and am now perhaps become over-cautious; and feel inclined to cavil at
speculation when the direct and immediate effect of a caus