303 



The figures for Ra 1, Ra 2, Ra 3, are for {3 rays only, the 

 effects of y rays having been eliminated. The figures for 

 Ra 4 have not been corrected for y rays, and must, indeed, 

 refer almost entirely to rays of the latter class. 



The velocity of the /8 rays of Ac is less than in the case 

 of UrX. In the case of radium it varies; some of the rays 

 are as little penetrating as those of Ac, others more than the 

 rays of UrX. The figures seem to show that there is a 

 certain small dependence on the velocity of the ^ rays, but 

 it is insufficient to account for the variations in quality which 

 have been described above. 



We must therefore fall back upon a second explanation, 

 viz., that the scattered rays possess velocities which are less, 

 on the average, than those of the primary ; and that the dif- 

 ference is greater in the case of the atoms of small atomic 

 weight. This is in general accordance with other experi- 

 ments and with expectation. It has been shown by Gehrcke 

 (Ann. der Phys., viii., p. 81, 1902) that scattered cathode 

 rays possess a variety of velocities, none greater than that of 

 the primary, some, however, being equal to it. And it is 

 not unreasonable to suppose that /3 rays lose more energy in 

 being scattered by atoms of less weight or of looser building 

 than by those which are heavier or more rigid. If, there- 

 fore, a homogeneous beam of j8 rays fell upon an aluminium 

 plate, there would be found in the scattered secondary rays 

 a larger proportion of particles of much reduced speed than 

 if the plate had been lead. 



When we measure the ionisation produced in an iouisa- 

 tion chamber, and insert successive thicknesses of absorbing 

 material between the radiating substance and the chamber, 

 placing them in contact' with the wall of the latter, the effect 

 is the same as if we gradually removed the chamber to greater 

 distances; if we may assume that the law of absorption in air 

 and in the given absorbing material is not greatly affected by 

 the alterations in quality which we are considering. We may, 

 therefore, take the figures in any vertical column of the first 

 table given above as showing the effects of the secondary 

 radiation at various distances from the radiator. If we plot 

 an ionisation-distance curve the whole area between the curve 

 and the axes is the true measure of the ionisation due to the 

 returned rays. The usual plan is to accept as a measure the 

 one figure at the head of the column. 



We find that in each case the curve may be nearly ex- 



- A .r 

 pressed as the sura of two exponentials as Ae ^' + 



- A ft' 

 Be ■^' ■ Whether this has a physical meaning or not does 



not concern our immediate argument. Commencing with 



one thickness of tinfoil on the dome, the actual figures are — 



