420 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 19 
initial density is difficult to determine. It has been placed all the 
way from 2.7 to 3:7 by various investigators. It is generally agreed 
that although the average density of surface rocks is from 2.7 to 
2.8, corresponding to granite or granodiorite, nevertheless the granitic 
layer is relatively few miles deep (say 5 to 20); and that underneath 
this very thin skin of granitic (and sedimentary) rocks lies a more 
basic material such as gabbro or even pyroxenite or peridotite. 
For the moment it will be sufficient to note in Fig. 2 the density 
curves with two initial densities, 3.0 and 3.5, corresponding respec- 
tively to average gabbro and to dense peridotite. Although the 
calculation was carried only to a depth of 3400 km, this limit being 
set by the seismologic data, it is clearly evident that the density is 
not increasing fast enough to make the mean density of the Earth 
equal to 5.5. For the two assumed surface densities the average 
density below 3400 km would be 15 and 20 respectively—obviously 
much too high to be reached by any reasonable extrapolation of 
the density curves. The high central density demanded by the 
density curves of Fig. 2 may be considered a consequence of the fact 
that the core of radius 3000 km has only 1/9 of the volume of the 
Earth whereas 0.3 to 0.4 of the mass remains to be accounted for. 
It is therefore impossible to explain the high density of the Earth 
on the basis of compressibility alone. The dense interior can not 
consist of ordinary rocks compressed to a small volume. We must 
therefore fall back upon the only reasonable alternative, namely, the 
presence of a heavier material, presumably some metal, which, to 
judge from its abundance in the Earth’s crust, in meteorites, and in 
the sun, is probably iron. We thus arrive at the conclusion accepted 
by the majority of geophysicists, but, in addition, we have here (1) 
a quantitative estimate of the increase of density due to compression 
alone, and (2) direct evidence of the presence in the See of the 
Earth of a dense material such as iron. 
Effect of temperature. This is a disturbing and uncertain factor. 
From the known temperature gradient at the surface it follows that 
the temperature at 100 km depth must be considerably above the 
melting-point of ordinary rocks; and it seems unlikely that the central 
temperature can be less than several thousand degrees. ‘The effect 
of this high temperature on the density is not easily estimated, and 
might conceivably be very large, but it so happens that the problem 
is simplified by the fact that at high pressures the expansion coefficient 
becomes less than at low pressures. Now, the pressure half way down 
to the center of the Earth is more than a million atmospheres, and 
