CHEMISTRY: IV. D. HARKINS 
223 
dance of these elements is due to changes which have taken place in the 
Kthosphere, or to the rising to the surface of the Ughter elements, but 
these objections are not so valid when the meteorites are found to show 
the same relations. The density of the earth's surface rock averages 
between 2.70 and 2.75, the mean density of the earth is 5.516, and the 
density of its center has been estimated by Lunn^ as 9.6 on the basis of 
Roche's law of density, and on the supposition that the chemical com- 
position of the earth is uniform. Stone meteorites vary in density from 
2.5 to 5, and iron meteorites from about 6 to more than 8, with an aver- 
age density of 7,8. According to Lunn^ the pressure at the center of the 
earth is 2,800,000 atmospheres, and a possible central temperature is 
16,610° when both are calculated on the basis of Roche's law, p = po 
(l—cx^). It seems probable that this law is much more in accord with 
the behavior of material than the simple Laplacian form usually used. 
Some writers have argued from the data that the center of the earth is 
mostly iron. However the extremely long range of extrapolation above 
the experimental values in both temperature and pressure, makes it 
seem impossible to get results in this connection which have the least 
value, however desirable it would be for such a problem as the one 
presented here if such a deduction could be properly made. Perhaps, 
then, the most that can be said is that in the three classes of material, 
the lithosphere, the stone meteorites, and the iron meteorites, in spite 
of variations in density from 2.5 to 8, the same two rules are found to 
hold, that (1) the even numbered elements, and (2) the elements of 
low atomic number and low atomic weight, are the elements which 
occur in abundance. 
If an artificial Hne of division is made just after the first eighth group 
in the periodic model so as to classify the first 29 elements as of low 
atomic number and atomic weight, and the remaining 63 elements as of 
high atomic weight, then the following table, based upon data from 
analyses listed by Farrington and Clarke may be presented to empha- 
size the importance of the former class. 
TABLE 3 
Proportion in Various Materials of the Elements of Low Atomic Numbers 
Percentage of Elements with 
Atomic Numbers 
Material 
1-29 
30-92 
Meteorites as a whole 
99.99 
0.01 
99.98 
0.02 
100.00 
0.0 
Igneous Rocks 
99.85 
0.15 
0.05 
0.05 
Lithosphere 
99.85 
0.15 
