208 
ME. EOBEET MALLET ON VOLCANIC ENEEGY. 
shell may pass into the nucleus if the latter be liquid ; it is only necessary for the author 
to postulate a higher temperature, and therefore a larger coefficient of contraction, for 
the interior of the globe than for the colder shell which surrounds it, and to suppose 
as was done by the late Mr. Hopkins in his researches as to the thickness of the shell 
4n relation to precession, that, whatever thickness may he assigned to the shell, it passes 
jper saltum into the nucleus — all that is here meant being, that all below this imaginary 
couche contracts more than does all above it for a given decrement of temperature of 
both. We have no certain knowledge of the rate at which temperature increases either 
in the shell or the nucleus in descending from the surface, nor what may be the highest 
temperature of the nucleus itself ; but as the mean temperature of the shell may he 
presumed greatly inferior to that of the nucleus, it may be allowable to regard the 
whole of the heat dissipated from our globe in a unit of time (a year) as derived from 
the nucleus only, and transmitted merely through the shell, the thickness of the latter 
being taken as not too large in relation to the earth’s radius. The total heat dissipated 
from our globe in a year, or, on the above suppositions, from the nucleus only, being, as 
above stated, equal to that evolved by the melting of 777 cubic miles of ice at 32° to 
water at 32°, may be considered for any moderate secular period, such as 5000 years, as 
constant. The refrigerative power of the unit of volume of a cubic foot of such ice is 
C=gX$, 
g being the specific gravity and s the latent heat of ice. Therefore 
C=57-8xl43=8265°-4 Fahr., 
or units of heat, assuming the capacity for heat of water to be the same at all tempe- 
ratures ; and the refrigerative effect of this upon an equal volume of the mass of the 
nucleus is 
s' and §' being the specific heat and specific gravity or weight per unit of volume, 
respectively, of the matter of the nucleus. We in reality know nothing as to what may 
be the chemical or physical nature of the matter composing the nucleus ; we therefore 
have no basis for assigning its specific heat in whole or in part ; nor do we know any thing 
as to its specific gravity beyond this, that the mean density of our globe being 5 - 5, that 
of the nucleus alone must be somewhat greater. We are therefore obliged to adopt 
the most probable suppositions we can for the values of s' and g'. It is highly probable, 
as appears to be generally conjectured by geologists, that a large proportion at least of the 
entire mass of our globe, and therefore of the nucleus as here defined, consists of rocky 
material not very dissimilar from that known to us by observation or inference in the 
superficial crust of the earth. Now as none of the materials of the crust, excluding 
those of metallic veins or beds relatively small in quantity, at all approach the average 
density of 5’5, we may reasonably conclude that towards the centre of our planet there 
exist masses of metals, the only bodies we are acquainted with whose high specific 
