Further data are required in order to fix which of the time constants of the changes refers to the first change. In order to settle this point, it is necessary to isolate one of the products of the changes and to examine the variation of its activity with time. If, for example, a product can be separated whose activity decays to half value in 55 minutes, it would show that the second change is the more rapid of the two. Now Pegram[302] has examined the radio-active products obtained by electrolysis of thorium solutions. The rates of decay of the active products depended upon conditions, but he found that, in several cases, rapidly decaying products were obtained whose activity fell to half value in about 1 hour. Allowing for the probability that the product examined was not completely isolated by the electrolysis, but contained also a trace of the other product, this result would indicate that the last change which gives rise to rays is the more rapid of the two. This point is very clearly brought out by some recent experiments of Miss Slater[303], who has made a detailed examination of the effect of temperature on the active deposit of thorium. A platinum wire was made active by exposure for a long interval to the thorium emanation, and then heated for a few minutes to any desired temperature by means of the electric current. The wire, while being heated, was surrounded by a lead cylinder in order that any matter driven off from it should be collected on its surface. The decay of activity both of the wire and of the lead cylinder was then tested separately. After heating to a dull red heat, no sensible diminution of the activity was observed at first, but the rate of decay of the activity on the wire was found to be more rapid than the normal. The activity of the lead cylinder was small at first but increased to a maximum after about 4 hours and then decayed at the normal rate with the time. These results are to be expected if some thorium A is volatilized from the wire; for the rise of activity on the lead cylinder is very similar to that observed on a wire exposed for a short time in the presence of the thorium emanation, _i.e._, under the condition that only thorium A is initially present. On heating the wire above 700° C. the activity was found to be reduced, showing that some thorium B had also been removed. By heating for a few minutes at about 1000° C. nearly all the thorium A was driven off. The activity on the wire then decayed exponentially with the time, falling to half value in about 1 hour. After heating for a minute at about 1200° C. all the activity was removed. These results show that thorium A is more volatile than B, and that the product which gives out rays, viz. thorium B, has a period of about 55 minutes.