™. 
330 Conjectures concerning the Cause, and Observations 
first, but by the great exteusion of the heated matter, yet, after 
they have once taken place, they may well continue te do so for 
some time; for the great disturbance in the first instance, by 
the falling in of a great part of the roof, must render the fre- 
quent communication between the fire and water not only very 
easy, but almost unavoidable: and this will continue to be so, 
till the roof is well settled, and the surface of the melted matter 
sufficiently cooled, after which, it may require a long time for 
the fire to heat it again so much, as will be necessary to make it 
produce the former effects. Now, as the matter has been more 
or less cooled, or as the combustible materials are with more or 
less difficulty set on fire again, as well as on account of other 
circumstances, the returns of these effects will be later or earlier; 
but though they will not, for this reason, observe any exact pe- 
riod, yet ‘they will generally fall within some sort of limits, till 
either the matter that occasions them is consumed, (which, pro- 
bably, will seldom happen in less than many ages,) or till the fires 
open themselves a passage, and become volcanos. 
Section IV,—77. I have already intimated, that the most 
extensive earthquakes frequently take their rise from the sea. 
According to the description of the structure * of the earth be- 
fore given, any combustible stratum must lie at greater depths 
in places under the ocean, than elsewhere; hence far more ex~- 
tensive fires may subsist there, than where the quantity of mat- 
ter over them is less; for any vapour raised from such fires, 
having both a stronger roof over it, and being pressed by a 
greater weight, (beside the additional weight of the water) will 
not only be less at liberty to expand itself, and consequently of 
less bulk, but it will also be easily driven away towards the parts 
round about, where the superincumbent matter is less, and there- 
fore lighter. On the other hand, any vapour raised from fires, 
where the superincumbent matter is lighter, fiuding a weaker roof 
over it, and being not so easily driven away under strata, that are 
thicker and heavier, will be very apt to break through, and open 
a mouth to a volcano; and it must necessarily do this long be- 
fore the fires can have spread themselves sufficiently, to be near 
equal to those which may subsist in places that lie. deeper. All 
this seems to be greatly confirmed by the situation of volcanos, 
which are almost always found on the tops of mountains +, and 
those often some of the highest in the world. 
78. If, 
* See art. 45. 
+ Perhaps this may supply us witha hint (if the conjecture is not thought 
extravagant) concerning the manner in which these mountains have been 
raised, and why the strata lie generally more inclining from the mountainous 
countries, than those countries themselves; an appearance not easily to be 
accounted for, but upon the supposition, that the upper parts of the earth 
rest 
