ON THE TIIEORIES OF ELEVATION AND EARTHQUAKES. 55 
the process begins ac the surface of the general mass), the effect of increased 
pressure at greater depths, in proniotiog solidification, must bo /ess tlian that 
of increased temperature in preventing it, while the formation of solid floors 
to the internal cavities, according to the above reasoning, requirta that the 
effect of increased pressure at greater depths in preventing the generation of 
elastic gases should be grecUer tlian that of increased tcioperaturo in pro¬ 
moting it. These assumptions may not be incoiuslsUmt with each other, 
though they may Im regarded, perhaps, so far opjiosed to each other as to 
diminish materially the degree ol' probability belonging to the theory which 
essentially involves them. The theory which supposes the solidification of 
the earth to have coinmence<l at tlie centre, assumes the efficiency of great 
pressure both in promoting solidificatioa and preventing the generation of 
elastic gases, and therefore involves no such possible inoonsUtency os that 
now indicated. 
IS. In the previous reasoning, ru» applied either to the case of the solidifi¬ 
cation beginning at the ceoire, or to that in which it begins at the surface, it 
is suppos^ tliat though the constant stute of motion in which the superficial 
particles of tlie generd fluid niasa were nmintnined by the ascent und escape 
of elastic gas^, ceased to be sufficient to preserve the fluidity at the surface 
at the time of the incipient superficial incrustation, still, that the same cause 
WAS sufficient at subsequent epochs to arrest, in a great degree, the process 
of solidification iimucdistely beneath the dislocated |wrtioJW of the crust. 
This may posdbly present itself as a difficulty; hut we may remark that the 
external temperature to which the external surface at the time of incipient 
incrustation would be exposed, would l>c very much lower than that to which, 
after the formation of a thin Buperticial crust, the portion of the fluid mass 
immediately beneath it would be exposed. Consequently a much less agi¬ 
tation of the fluid particles among lliomselvcs produced by ihe m<cent of 
vapours, would suffice to arrest the process of solidification ihnu Uiat which 
would be necessary to prevent its conimencemeiit. Again, previously to the 
superficial incrustation, the elastic vapours would have an w|ually free egress 
from every point of the general surface, wherems, after the formation of a 
crust, this egnss would bo restricted to particular localities to which the 
ascending streams of vapours would converge in their ascent, und thus be¬ 
come by thoic^nion more ollective in aiTCF>ting solidifleatiou in those locali- 
ties, as already explained. At later periods, also, auoUicr change in the 
conditions of the muss would take place, which might he of coosidvrablc im¬ 
portance. When the solidification of the surfaco began, the. whole of the 
present fluid matter on tlic globe must havee.xisted in n gaseous state in tbc 
form of an atmosphere, produinng at every point of tlie surface a pressure 
euonnoua aa compared with the present atiuosphuric presaiire. Allowing, 
tlierefore, the influence of pressure in preventing tho generation of elastic 
vapours, it U very possible that incnistatiou might iuko place at a much 
higher temperatuiv! than that wlitcli is now necu-ssAry to preserve fluidity in 
extsting volcanic niiwses. 
Another cause may possibly have- been effective aUu in preserving the 
ma^ea of fluid mailer wlilch constitute existing volcanos. It would appear 
to be not improbable, that tlie more fusible portions of the general terrestrial 
ma^ vrould have, on the whole, a smaller specific gravity in their state of 
miou than those portions which were less fusible, in which caiw there would 
w a nect'f«ary tendency to thq aixumulation of tho more fusible parts under 
those portions of the solid crust beneath which the ascent of clastio vapours 
was most active and longest continued. E.xistiug volcanos may possibly be 
due in some degree to such accumulations. 
