Royal Institution. 469 
for heat is communicable, and the surface in the course of 
ages would have gained the same temperature. 
We can reason only by analogy, from known concerning 
unknown phenomena. Fire on the surface of the globe 
is usually a result of chemical changes: it is therefere 
reasonable to infer that subterranean fires depend upon si- 
milar causes ; and the idea is supported by their cessation, 
renovation, and varied duration. 
But what are the agents concerned in these great and 
awful elevations? The discoveries of Mr. Davy prove that 
the earths and alkalies consist of metals united to oxygen, 
or pure air; and these metals are highly inflammable, some 
of them so much so as to burn even in contact with water. 
The mean density of the earth, as determined by Mr. Ca~ 
vendish and Dr. Maskelyne, would lead to the conclusion 
that the interior. consists principally” of metallic matter, 
which may be alloys of the metals of the earths and alkalies 
with the common metals :—and such an assumption, says 
Mr. Davy, would offer a ready explanation of subterranean’ 
heat and. volcanic explosions; for, supposing water from 
the sea or Jakes to act upon these inflammable masses, elastic 
matters would be rapidly disengaged, the surface would 
be broken, air would)act upon the metals, inflammation 
would take place, andothe result’ would be lava, the metals 
of the earths combined-withsoxygen. . 
Even.a. general hypothesis ain Geology might be con- 
nected with. theisamesideas eRoekwrtaresdecomposed and 
degraded-by water; -assisted:by heatrard atmospheric electri- 
city. By othe operations; afsvaléanesy lad’ ie raised and 
matter newly. consolidateds; sand itis possible ‘to conceive 
electrical currentstinestbenglobey by Which ietallic matter 
may, be separated fromm yeeny 90/485 tO/preserve a perfect 
equilibrium betivetiicalhthe partsbofohe/system,—an order 
in nature which Mmaybe-conceived tosbé represented by the 
ancient-hieroglyphio: ofithe:Phositixirisitig, from! her ashes. 
The reasoning derivéd:frothothe:eonsideration of vol- 
canic fires relate to the faturerorder of! things) but cannot! 
with propriety be appli¢th taitheoformation of “primary and 
secondary rocks. ,Asvyetsn lavasthave/ beer found ana- 
logous to granite or porphyry ;iandithough difference of 
pressure may interfere, yet this should not be assumed, but 
proved by experiments: | ‘ 
The obvious and immediate effect of volcanic: fires 1s to 
increase the extent of the surface of the globe, and to ratse 
land from the sea, and the soils produced from the decom- 
position are generally very fertile-—Some of the most beau- 
Geg3 tifa} 
