After the heating effect of the emanation tube had attained a maximum, the emanation was removed, and the decay with time observed as soon as possible afterwards. The results are shown in the lower curve of Fig. 101. It is seen that the two curves of rise and decay are complementary to one another. The first observation was made 10 minutes after removal, and the heating effect had then dropped to 47 per cent. of the original value. This sudden drop is due partly to the removal of the emanation, and partly to the rapid transformation of radium A. The lower curve is almost identical in shape with the corresponding α ray curve for the decay of the excited activity after a long exposure (see Fig. 86) and clearly shows that the heating effect is directly proportional to the activity measured by the α rays over the whole range examined. The heating effect decreases according to the same law and at the same rate as the activity measured by the α rays.
Twenty minutes after the removal of the emanation, radium A has been almost completely transformed, and the activity is then proportional to the amount of radium C present, since the intermediate product B does not give out rays. The close agreement of the activity and heat emission curves shows that the heating effect is proportional also to the amount of radium C. We may thus conclude that the rayless product B supplies little if any of the heat emission observed. If radium B supplied the same amount as radium C, the curve of decrease of heating effect with time would differ considerably from the activity curve.
The conclusion that the transformation of radium B is not accompanied by the release of as much heat as the other changes is to be expected if the heating effect is mainly due to the energy of motion of the expelled α particles.