﻿Absorption and Scattering of the ex. Hays. 919 



results, the values of the most probable angles of scattering 

 deduced from the numbers in the six columns of Table I. 

 For the thick foils (13) was used, for the thin (12). I have 

 calculated the latter for both 1 and 2 foils, as Geiger's 

 measurement with one foil seemed to be in not very good 

 agreement with the next few. 



Except fcr the last column the results are of the right 

 order. The divergence of the last column is chiefly due to 

 the fact that log iv 2 — a 2 does not then in the least approxi- 

 mate to log (I + W2)— ; f 2 (ic 2 ). In the case of gold indeed 

 log w 2 — a 2 changes sign. In spite of this divergency for 

 the surface distribution expressions, which masks their true 

 value, it does seem possible now to discriminate definitely in 

 favour of the volume distribution, and it seems that the 

 smaller values of a are best. Since the closeness of the 

 approximation of % to its true value is unknown, it does not 

 seem profitable to fix the value with any greater accuracy. 



The agreement between the observed and calculated values 

 for the scattering is thus good enough entirely to confirm 

 the hypothesis of this paper, but not so good as to help much 

 in a more accurate specification of the atomic constants. 



Summary. 



An hypothesis is put forward whereby the a particles in 

 passing through matter pull electrons out of the atoms they 

 traverse, acting on them with the ordinary law of the inverse 

 square. 



An equation is deduced relating iheir velocity to the 

 distance they have travelled from their source. This is 

 the " velocity curve " and agrees closely with the experi- 

 mental curve. 



The equation involves two unknown constants : n the 

 number of electrons in each atom, a the radius of the atom. 

 In the case of air, if a be assumed known, n can be deduced 

 from the range. Widely different values of cr give very 

 similar values of n. 



From comparison of the stopping powers of air and other 

 substances, a and n for these can be deduced, the values all 

 depending on the original value assumed for the radius of 

 the air atom. 



The number of electrons in the atom appears to be inter- 

 mediate between the atomic weight and its half. The atomic 

 radii decrease with increasing atomic weight. 



In the case of hydrogen it seems probable that the formula 

 for a does not hold on account of there being only very few 



