480 Prof. Bumstead and Mr. McGougan on Emission of 



atoms) may make five times as many ions as the a-rays. 

 In our experiments the current of S-electrons produced by 

 the soft primary radiation is only J of that produced by the 

 a-rays. This discrepancy is due, at least in part, to the fact 

 that the deposit of polonium which we used could not be 

 regarded as " thin " in regard to this very absorbable radiation. 

 Wertenstein found that this radiation could be considerably 

 increased by depositing the Ra G upon a platinum plate, and 

 thus avoiding a film of oxide which formed on other 

 metals under the action of the radium emanation. Another 

 cause for the difference is probably to be found in the fact 

 that his experiments were made at pressures of from 3 to 

 10 mm. instead of in a high vacuum ; in the latter case, 

 fewer of the recoil atoms are charged, and their ionizing 

 effect might be expected to be less. 



On the other hand, the ionizing effect of the secondary 

 rays in his experiments was small, amounting only to about 

 J of the a-ray effect, whereas our results show that it causes 

 the emission of 1*12 times as many S-electrons as the a-rays. 

 It is true that these results depend upon the dimensions of 

 the cylinders which are used to limit the pencil of a-rays, 

 but it seems scarcely possible that so large a discrepancy can 

 be attributed to this cause. It may be that the secondary 

 rays are more efficient in causing the liberation of S-electrons 

 from the metals than in ionizing a gas ; it may be also that 

 the transverse magnetic field used by Wertenstein to deflect 

 these rays from his ionization-chamber was not altogether 

 effective in the presence of air which was necessary for 

 ionization measurements. 



The values of s in the second column of Table IV. give 

 some idea of the distribution of velocities among the secondary 

 electrons, but no numerical results can be conveniently 

 deduced from them. The number of secondary (in this case 

 tertiary) electrons emitted by a metal when struck by a 

 moving electron changes rapidly with the speed of the 

 moving electron, especially within the range of velocities with 

 which we have to do in this case *. It will probably not be 

 difficult, however, to determine the velocity distribution by 

 using a Faraday cylinder as the recipient of the rays. 



It is evident that the range of velocities extends from less 



cm 

 than 3'8 x 10 8 — -, which corresponds to 40 volts, to more 



sec. l 



than 18"8xl0 8 , which corresponds to 1000 volts. An 

 experiment, not here reported in detail, showed that there 



* See Gehrts, Ann. d. Phys. xxxvi. pp. 1003 ct scq. (1911). 



