When a polished bismuth disk was kept for some hours in the solution, it became strongly active. The active matter deposited on the bismuth gave out α rays, but no trace of β rays. After several bismuth disks had been successively left in the solution, the active matter, which emits α rays, was almost completely removed. This was shown by evaporating down the solution after treatment. The β ray activity remained unchanged, but that of the α rays had been reduced to about 10 per cent. of its original value. Three bismuth disks, made active in this way, were set aside and their activity measured at regular intervals. The activity fell off according to an exponential law with the time during the 200 days since their removal, while that of each fell to half value on an average in about 143 days. At the same time it was observed that the solution, from which the α ray activity was removed, gradually regained its activity, showing that the active substance which gave out α rays was continuously produced from the matter left behind in the solution. =234. Explanation of the results.= We have seen that a close examination of the active deposit of slow change has disclosed, (1) the presence of a β ray product which loses half of its activity in about six days; (2) the presence of an α ray product, which is deposited on bismuth and is volatilized at 1000° C. This product loses half of its activity in 143 days; (3) the presence of a parent substance, which produces the β ray product at a constant rate. This parent product must be transformed very slowly since the β ray product, which arises from it, soon reaches an equilibrium value, which does not change appreciably over a period of more than one year. The experimental evidence points to the conclusion that the parent product does not give rise to β rays, but that the β rays arise entirely from the next product. This parent product cannot give rise to α rays, for we have seen that the initial α ray activity is at first extremely small, but increases steadily with the time for a period of at least eighteen months. Thus the parent product does not give rise to either α or β rays, and must be a “rayless” product. The first three transition products of the radium emanation, viz. radium A, B and C, have already been analysed, and shown to be consecutive. It thus seems probable that the active deposit of slow change must arise from the successive transformations of the last product radium C. The results already obtained can be completely explained if it is supposed that three transition products, viz. radium D, E and F, are present in the active deposit of slow rate of change. The properties of these products are summarized below. _Radium D_ is a rayless product of very slow rate of change. It will be shown later that it is half transformed in about 40 years. It is volatile below 1000° C. and is soluble in strong acids.