evel THE PRESIDENTIAL ADDRESS. 
peculiarities which are at all events of considerable interest. I may add 
that it is contemplated that the Cambridge apparatus should also be 
designed to eliminate the disturbing element I have mentioned, and that 
it should be available for determinations at sea. It is perhaps not too 
much to hope that with the co-operation of the Navy, the gravity chart of 
the world, which is so far almost a blank as regards the ocean, may in 
this way be gradually filled in. 
The distribution of the intensity of gravity over the surface of the earth 
gives by itself no positive information as to the distribution of density 
throughout the interior, though the contrary view has sometimes been held. 
For example, a spherical globe with a uniform intensity of gravitation over 
its surface would not necessarily be homogeneous, or even composed of 
spherical strata each of uniform density, however plausible this might be 
on other grounds. Consequently, there is room for hypothesis. There are 
certain tests which any hypothesis has to satisfy. It must account for the 
observed distribution of gravity, and having regard to the phenomena of 
precession, it must give the proper relation between the earth’s moments 
of inertia about a polar and an equatorial axis. It may be added that it 
should be fairly consistent with the ascertained velocities of seismic waves 
at different depths, and the degree of elasticity which it is allowable to 
assign to the material. The somewhat artificial laws of density adopted 
by Laplace and Roche, respectively, mainly on grounds of mathematical 
convenience, have lost much of their credit. A more natural law, suggested 
indeed by Thomson and Tait in 1867 in their book on Natural Philosophy, 
has since been proposed in a more definite form by Wiechert. On this view 
the earth is made up of a central core of about four-fifths the external 
radius, of high density, about that of iron, surrounded by an envelope of 
about the density of the surface rocks. This is, of course, only to be taken 
as a rough picture, but it satisfies the requirements I have mentioned, and 
is apparently not incompatible with the seismic data. 
In all speculations on the present subject, considerations as to the 
thermal history of the earth and the present distribution of temperature 
in the interior play an essential part. The apparent inconsistency between 
the requirements of physics and geology was long a matter of controversy, 
and has given rise to keen debate at these meetings. Lord Kelvin’s historic 
attempts to limit the age of the earth by consideration of the observed 
temperature gradient as we go downwards from the surface lost their 
basis when it was discovered that the rate of generation of heat in the 
processes of radioactive change was amply sufficient to account for the 
