$0 Scientisic Proceedings, Royal Dublin Society. 
circumstance just referred to makes it now more than usually 
allowable to argue from experiments under limited ranges of 
condition. Sir William declares that he believes it likely that 
the materials of the globe, though on the whole as rigid as 
steel, may be yet, at every depth, not much below the melting 
temperature corresponding to the pressure at that depth. Ifthis 
should be so it would tend to make the effects of the increasing 
temperature and of the increasing pressure to be generally com- 
parable with each other. Though the two influences concerned 
are each of high intensity, according to our ideas, as they are 
opposed, and perhaps somewhat near a balance in a certain respect, 
they may produce, as regards this, a mean result which is probably 
accordant with physical laws, in forms not very different from 
those manifested by them under the nearest laboratory conditions. 
To this we may add that the critical point of fusion and the 
consideration of its latent heat is happily avoided. But for these 
two circumstances it would be hardly worth while to enter on the 
following argument. 
Within the range of experiment the general law seems to be 
that, with rise of temperature, the absolute conductivities of 
substances diminish,* and their thermal capacities increase, so that, 
for a two-fold reason, when the temperature of a body is raised its 
power to transmit diminishing heat becomes lessened. On the 
other hand increase of pressure, as a general rule, acts in the opposite 
way ; it causes the absolute conductivities of substances to increase, 
and their thermal capacities to diminish. But before proceeding 
further let us note that, with metals at least, even if the con- 
densation due to high pressure be equal to the expansion due 
to increased intensity of heat the thermal conductivity will still 
be diminished. So that if the density of such substances remains 
unaltered, the high temperature, though only just a match for the 
high pressure, as far as its effect on density is concerned, is more 
than a match for the high pressure in diminishing the heat-trans- 
mitting power of such substances. 
Now let us note a circumstance that adds to the importance in 
the present question of elevation of temperature. The rate of 
expansion of heated bodies increases with the rise of temperature; 
this goes to promote the diminution of the thermal conductivity 
* See note at end. 
