source of radiation. The radium, sealed in a glass vessel, was surrounded by an external glass tube and exposed to the action of the β and γ rays of radium for several weeks. The inside and outside of the glass tube were found to be active, and the active matter was removed by solution in water. The radio-activity observed was very minute, corresponding to only about 1 milligram of uranium. The writer has, at various times, tried experiments of this character but with negative results. The greatest care is necessary in such experiments to ensure that the radio-activity is not due to other causes besides the rays from the radium. This care is especially necessary in laboratories where considerable quantities of the radium emanation have been allowed to escape into the air. The surface of every substance becomes coated with the slow transformation products of radium, viz. radium D, E, and F. The activity communicated in this way to originally inactive matter is often considerable. This infection by the radium emanation extends throughout the whole laboratory, on account of the distribution of the emanation by convection and diffusion. For example, Eve30 found that every substance which he examined in the laboratory of the writer showed much greater activity than the normal. In this case the radium had been in use in the building for about two years.
265. Loss of weight of the radio-elements. Since the radio-elements are continually throwing off α particles atomic in size, an active substance, enclosed in a vessel sufficiently thin to allow the α particles to escape, must gradually lose in weight. This loss of weight will be small under ordinary conditions, since the greater proportion of the α rays produced are absorbed in the mass of the substance. If a very thin layer of a radium compound were spread on a very thin sheet of substance, which did not appreciably absorb the α particles, a loss of weight due to the expulsion of α particles might be detectable. Since e/m = 6 × 103 for the α particle and e = 1·1 × 10-20 electromagnetic units and 2·5 × 1011 α particles are expelled per second per gram of radium, the proportion of the mass expelled is 4·8 × 10-13 per second and 10-5 per year. There is one condition, however, under which the radium should lose in weight fairly rapidly. If a current of air is slowly passed over a radium solution, the emanation produced would be removed as fast as it was formed. Since the atom of the emanation has a mass probably not much smaller than the radium atom, the fraction of the mass removed per year should be nearly equal to the fraction of the radium which changes per year, i.e. one gram of radium should diminish in weight about half a milligram (section 261) per year.
If it is supposed that the β particles have weight, the loss of weight due to their expulsion is very small compared with that due to the emission of α particles. The writer has shown (section 253) that about 7 × 1010 β particles are projected per second from 1 gram of radium. The consequent loss of weight would only be about 10-9 grams per year.
Except under very special experimental conditions, it would thus be difficult to detect the loss of weight of radium due to the expulsion of β particles from its mass. There is, however, a possibility that radium might change in weight even though none of the radio-active products were allowed to escape. For example, if the view is taken that gravitation is the result of