one ten-millionth part of the attraction of the earth. The
attraction of such a model has actually been measured. Since we do
not know the average specific gravity of the earth--that being in
fact what we want to find out--we take a globe of lead, four feet
in diameter, let us suppose. By means of a balance of the most
exquisite construction it is found that such a globe does exert a
minute attraction on small bodies around it, and that this
attraction is a little more than the ten-millionth part of that of
the earth. This shows that the specific gravity of the lead is a
little greater than that of the average of the whole earth. All
the minute calculations made, it is found that the earth, in order
to attract with the force it does, must be about five and one-half
times as heavy as its bulk of water, or perhaps a little more.
Different experimenters find different results; the best between
5.5 and 5.6, so that 5.5 is, perhaps, as near the number as we can
now get. This is much more than the average specific gravity of
the materials which compose that part of the earth which we can
reach by digging mines. The difference arises from the fact that,
at the depth of many miles, the matter composing the earth is
compressed into a smaller space by the enormous weight of the
portions lying above it. Thus, at the depth of 1000 miles, the
pressure on every cubic inch is more than 2000 tons, a weight
which would greatly condense the hardest metal.

We come now to the planets. I have said that the mass or weight of
a heavenly body is determined by its attraction on some other
body. There are two ways in which the attraction of a planet may
be measured. One is by its attraction on the planets next to it.
If these bodies did not attract one another at all, but only moved
under the influence of the sun, they would move in orbits having