ds. Even gases may be expected to some extent to be
retained by occlusion. Among the contents of uranium minerals, then, we
may look for the descendants of the parent uranium. If the descendants
are permanent or more long-lived than uranium, they will accumulate
continually. If they are short-lived, they will accumulate at a steady
rate till enough is formed for the quantity disintegrating to be equal
to the quantity developed. A state of mobile equilibrium will then
be reached, and the amount of the product will remain constant. This
constant amount of substance will depend only on the amount of uranium
which is its source, and, for different minerals, if all the product
is retained, the quantity of the product will be proportional to the
quantity of uranium. In a series of analyses of uranium minerals,
therefore, we ought to be able to pick out its more short-lived
descendants by seeking for instances of such proportionality.

Now radium itself is a constituent of uranium minerals, and two series
of experiments by R.J. Strutt and B.B. Boltwood have shown that the
content of radium, as measured by the radio-activity of the emanation,
is directly proportional to the content of uranium. (Strutt, "Proc.
Roy. Soc." A, February 1905; Boltwood, "Phil. Mag." April, 1905.) In
Boltwood's investigation, some twenty minerals, with amounts of uranium
varying from that in a specimen of uraninite with 74.65 per cent.,
to that in a monazite with 0.30 per cent., gave a ratio of uranium to
radium, constant within about one part in ten.

The conclusion is irresistible that radium is a descendant of uranium,
though whether uranium is its parent or a more remote ancestor requires
further investigation by the radio-active genealogist. On the hypothesis
of direct parentage, it is easy to calculate that the amount of radium
produced in a month by a kilogramme of a uranium salt would be enough
to be detected easily by the radio-activity of its emanation. The
investigation has been attempted by several observers, and the results,
especially those of a careful experiment of Boltwood, show that from
purified uranium salts the growth of radium, if appreciable at all, is
much less than would be found if the radium was the first product of
change of the uranium. It is necessary, therefore, to look for one or
more intermediate substances.

While working in 1899 with the uranium residues used by M. and Mme
Curie for the preparation of radium, Debie