ay be comparable with that of other
radio-active bodies, the relation between the amounts under
consideration must be a definite one. For this purpose equal weights
of the bodies are not taken, but use is made of the amounts which are
in equilibrium with a fixed amount of the parent substance.

One gram of radium has been settled upon as the standard for that
series and a unit known as the "curie" has been adopted to express the
equilibrium quantity of radium emanation. Thus, a curie of radium
emanation (or niton) is the weight (or, as this is a gas, the volume
at standard pressure and temperature) of the emanation in equilibrium
with one gram of radium. This, by calculation and experiment, is found
to be 0.63 cubic millimeter. When this amount has been produced by one
gram of radium, the formation and decay will exactly balance one
another. This is, therefore, one curie of emanation.

The measurement of the rate of decay is difficult but can be carried
out with great accuracy, even down to seconds, in the case of certain
short-lived bodies. Errors crept in at first from the failure to
completely separate the substances produced in the series, and
sometimes because of the simultaneous production of two substances.

As stated, the decay follows an exponential law. The time required for
the decay of activity to half-value does not mean, therefore, that
there will be total decay in twice that time. Thus the half-value
period for uranium _X_ is about 22 days. The period for complete decay
is about 160 days. This half-value period corresponds to the
half-value recovery period of uranium, which is also 22 days.

These were the earlier figures obtained for uranium _X_ and they
illustrate some of the difficulties surrounding such determinations.
It was found later that the body examined as uranium _X_ was really a
constant mixture and of course the decay and recovery periods were
also composite. It required later and very skilful work to separate
them into the bodies indicated in the disintegration series.

The half-value period for thorium _X_ is much shorter, namely, a
little over four days, and this is also the recovery period for
thorium _X_. The plotted decay and recovery curves will intersect at
this point.

The consecutive disintegration series, with the half-value periods,
for the uranium and thorium series as given by Soddy are seen in the
following tables. They are probably subject to some changes on further
an