When a radium salt has lost part of its activity by heating, the diminution is not lasting; the activity of the salt is spontaneously regenerated at the ordinary temperature, and approaches a certain limiting value. I have observed the curious fact that this limit is higher than the limiting activity of the salt before being heated—this, at least, is the case with the chloride. I give examples of this:—A specimen of barium-radium chloride which, after having been prepared in the solid state, has long since attained its limiting activity, possesses a total radiation represented by the number 470, and a radiation capable of traversing 0·01 m.m. of aluminium, represented by the number 157. This, specimen is heated to redness for several hours. Two months after the heating it attains its limit of activity with a total radiation equal to 690, and a radiation through 0·01 m.m. of aluminium, equal to 227. The total radiation and the radiation transmitted by aluminium are therefore increased respectively in the ratios 690/470 and 227/156. These two ratios are practically equal to one another, and are equivalent to 1·45. A specimen of radium-barium chloride which, after having been prepared in the solid state, has attained a limiting activity of 62, is maintained in a state of fusion for some hours; the fused product is then powdered. The product regains a new limiting activity equal to 140, which is twice as great as that to which it was able to attain when prepared in the solid state without having been sensibly heated during evaporation. I have investigated the law of increase of activity of radium compounds after heating. The following are the results of two series of determinations:—The figures of Table I. and II. represent the intensity of the radiation (I) as a function of time, the limiting intensity being supposed equal to 100, and the time being reckoned from the close of the heating. Table I. (Fig. 13, Curve I.) refers to the total radiation of a specimen of barium-radium chloride. Table II. (Fig. 13, Curve II.) relates to the penetrating radiation of a specimen of barium-radium sulphate, the intensity of the radiation which traversed 3 c.m. of air and 0·01 m.m. of aluminium having been determined. The two products were subjected to a bright red heat for seven hours.