By suitably choosing the conditions, the recovery curve can be made to rise normally from a constant minimum, and the decay curve be shown to consist of two curves, the first the rate of production of excited radioactivity, and the second the rate of decay of the activity as a whole. It is a significant fact that exactly similar curves have already been obtained by one of us (Phys. Zeit., 1902, 3, 254) for the excited radioactivity produced by the thorium emanation under very similar conditions. If a negatively charged wire be exposed for a few minutes only to the thorium emanation, the excited radioactivity produced at first increases to several times its value for the first few hours after the exciting cause is removed, and then commences to decay, exactly as in the case of ThX.

So far nothing has been stated as to whether the excited radioactivity which contributes about 21 per cent. of the total activity of thorium is the same or different from the known type produced by the thorium emanation. All that has been assumed is that it should follow the same general law, that is, the effect should increase with the time of action of the exciting cause, and decrease with time after the cause is removed. If the rate of increase of the excited activity be worked out from the curves given (Fig. 5) it will be found to agree with that of the ordinary excited activity, that is, it rises to half the value in about 12 hours. Curve I is the observed decay curve for ThX, curve II is the theoretical curve, assuming that it decreases geometrically with the time and falls to half value in four days. Curve III is obtained by plotting the difference between these two, and therefore constitutes the curve of excited activity. Curve IV is the experimental curve obtained for the rise of the excited radioactivity from the thorium emanation when the exciting cause is constant. But the exciting cause (ThX) in the present case is not constant, but is itself falling to half value in 4 days, and hence the difference curve, at first almost on the other, drops away from it as time goes on, and finally decays to zero. Curve III, Fig. 1, represents a similar difference curve for the decay of excited activity, plotted from the recovery curve of thorium. There is thus no reason to doubt that the effect is the same as that produced by the thorium emanation, which is itself a secondary effect of ThX.

IX. The Non-separable Radioactivity of Thorium.

It has not yet been found possible by any means to free thorium from its residual activity, and the place of this part in the scheme of radioactivity of thorium remains to be considered. Disregarding the view that it is a separate phenomenon and not connected with the major part of the activity, two hypotheses can be brought forward

capable of experimental test, and in accordance with the views advanced on the nature of radioactivity, to account for the existence of this part. First, if there was a second type of excited activity produced by ThX, similar to that known, but with a very slow rate of decay, it would account for the existence of the non-separable activity. If this is true, it will not be found possible to free thorium from this activity by chemical means, but the continuous removal of ThX over a very long period would, as in the above case, cause its spontaneous decay.

Secondly, if the change which gives rise to ThX produces a second type of matter at the same time, that is, if it is of the type of a decomposition rather than a depolymerisation, the second type would also in all probability be radioactive, and would cause the residual activity. On this view, the second type of matter should also be amenable to separation by chemical means, although it is certain from the failure of the methods already tried that it resembles thorium much more closely than ThX. But until it is separated from the thorium producing it, its activity will not decay spontaneously. Thus

what has already been shown to hold for ThX will be true for the second constituent if methods are found to remove it from the thorium.

It is shown in the following communication by one of us (p. 860) that uranium also possesses a non-separable radioactivity extremely analogous to that possessed by thorium, and whatever view is taken of the one will in all probability hold also for the other. This consideration makes the second hypothesis that the residual activity is caused by a second non-thorium type of matter produced in the original change the more probable of the two.

X. The Nature of the Radiations from Thorium and ThX.

It has recently been found (Rutherford and Grier, Phys. Zeit., 1902, 3, 385) that thorium compounds, in addition to a type of easily absorbed Röntgen rays, non-deviable in the magnetic field, emit also rays of a very penetrating character deviable in the magnetic field. The latter are therefore similar to cathode rays, which are known to consist of material particles travelling with a velocity approaching that of light. But thorium, in comparison with uranium and radium, emits a much smaller proportion of deviable radiation.

From the view of radioactivity put forward, it necessarily follows that the total radioactivity of thorium is altered neither in character nor amount by chemical treatment. This conclusion can be tested by comparing the radiations of thorium and ThX with the mixture which constitutes the thorium radiation. It must, however, be pointed out that it is difficult to make any absolute measurement of radioactivity on account of the different extents in different cases to which the radiations are absorbed in the material of the substance emitting them. The total radioactivity of the original thorium is derived from a small quantity of the substance in the form of powder, whilst the radiations from ThX are produced by a very thin film of the material on the platinum dish. The radiation from thorium is absorbed to the extent of one-half by aluminium foil 0.0004 cm. thick, and as thorium oxide is far denser than aluminium, it is probable that the radiation in this case is confined to a surface layer only 0.0001 cm. deep. In the ThX, on the other hand, there is probably but little absorbed in the substance itself. The difficulty can be overcome to some extent by taking for the comparison the radioactivity of a thin film of a soluble thorium salt produced by evaporating a solution to dryness over a large metal plate. Compared in this way, the radioactivity of ThX when first separated is almost exactly equal to the activity of the nitrate from which it is produced, whilst the hydroxide retains about two-fifths of this amount. The difference is in the

expected direction, for it is certain that more absorption takes place in the nitrate than in the products into which it is separated. The requirements of the hypothesis can thus be said to be satisfied, but the example illustrates the difficulty of making absolute measurements of radioactivity. These throughout have almost completely been avoided. It is possible to trace with great accuracy the change of radioactivity of any preparation by leaving it undisturbed on its original plate and comparing it always with the same comparison sample; but to express the radioactivity of one body like ThX in terms of that of another like thoria, except for the purposes of comparison, is misleading, as the above consideration shows.

Similar difficulties stand in the way of an answer to the second question, whether the nature of the radiations is affected by chemical treatment, for it has been observed experimentally that the penetrative power of these radiations decreases with the thickness of material traversed. The character of the radiations from ThX and thorium have, however, been compared by the method of penetrative power. A large number of comparisons justifies the view that the character of thorium radioactivity is unaltered by chemical treatment and the separation of ThX, although the different types are unequally distributed among the separated products.

The determination of the proportion between the deviable and nondeviable rays affords a new means of approaching the question. The general result is that the radiations from ThX and the excited radiation it produces both comprise deviable and non-deviable radiation. But in the experiment in which the excited radiation was allowed to decay spontaneously by removing the ThX as formed, the final product, after 23 precipitations, was found to be quite free from deviable radiation. This, as will be shown in the following paper, is one of the most striking resemblances between the non-separable radioactivities of uranium and thorium.

Finally, it may be mentioned that the proportion of deviable and non-deviable radiation is different for different compounds of thorium. The nitrate and ignited oxide, compounds which hardly possess any emanating power, have a higher proportion of deviable radiation than compounds with great emanating power. This is indirect evidence of the correctness of the view already put forward (section VII), that when the emanation is prevented from escaping it augments the proportion of excited radioactivity of the compound.

XI. Summary of Results.

The foregoing experimental results may be briefly summarised. The major part of the radioactivity of thorium-ordinarily about