140 
CLARKE. 
On the theoretical side the results attained are not easy to 
interpret. That nine of the chemical elements should con¬ 
stitute, at the lowest estimate, 98 per cent, of all known ter¬ 
restrial matter is somewhat startling and difficult to compre¬ 
hend. Are the other elements concentrated in the interior 
of our planet ? On this point there is a little positive evi¬ 
dence. 
The mean density of the earth, 5’5 to 5’6, is more than 
double that of the rocky crust, and the difference may be 
accounted for as a result of pressure, or by supposing that as 
the globe cooled the heavier elements accumulated toward 
the centre. Both suppositions may be true in part, but less 
weight is to be placed upon the second, for the following 
reason: A mixture of the elements in equal proportions, in 
the free state and as they behave at the earth’s surface, would 
have a specific gravity of about 7'3. In combination the 
density would be greater because of condensation, and below 
the surface it would also be increased by pressure. Hence it 
seems clear, since the density of the earth is only 5‘5, that in 
the planet as a whole the lighter elements must very consid¬ 
erably exceed in quantity the heavier ones. Twenty-nine of 
the known elements have densities below 5*5, and forty exceed 
that figure, iron being the only one of the heavier group which 
is at all abundant. The greater part of the earth’s mass is 
almost certainly to be found among the twenty-nine lighter 
elements. The others may be more plentiful at the centre 
of the globe than upon its surface, but few beside iron can 
be dominant constituents. This evidence seems to be clear, 
even though it is not proof positive. 
An attempt was made in the course of this investigation to 
represent the relative abundance of the elements by a curve, 
taking their atomic weights for one set of ordinates. It was 
hoped that some sort of periodicity might be evident, but no 
such regularity appeared. No definite connection with the 
periodic law seemed to be traceable. And yet certain other 
regularities are worth noticing. All of the abundant elements 
are low in the scale of atomic weights, reaching a maximum 
