( 352 ) 



27r TT - 



f .... da = a' i l . . . . (h] m> i)-(Jd- — 2 rr a' i . 



o (» 



R- 



ƒ■ 



s;ni>dt>: 



o 



R + n 



r dr 



2.a ; .... ^ 



R — a\ 



SO : 



-\- co fl + a 



4^^ = ^ f'^ (Vo-r(^-O)^^'' (9). 



2a J J^ J 



— (x> I R^n i 



The lower limit \R — (f] equals R — rr, if W^a, hnt equals a — R, 

 if R <^ (i- Now as follows from ihe definition of h\ the integral 

 with respect lo y is equal lo J or 0, according as the argument 



y c {t — t) vanishes for one of the values of y, contained between 



1 7^ — ,{\ and R-\-'i, or not. Wi-ile for ahhreviation : 



T = t — f', 



then our iidegral becomes equal to 1, if 



]i -f a ^ c r and \R — a, <^c t 

 that is, if ^ve can form a triangle with the sides I*,ii and r t : it 

 becomes equal to in the contrary case. If ?. means the nuuibei- 

 1 or 0. according as that triangle is possible or not, we may also wi-ite 



00 



8c f'Xdr 

 ^ 2a./ R 



instead of (9) ; reversing the lindts and noting that the cou(btion 

 \E — a\<^<'-r is never satisfied for r <^ 0. we may finally put: 



ao 



£ c r ). dx 

 4::t(p = — — — (10) 







This is exactly the formula [11) of my first paper. There is no 

 question of imagining a deformed shape of electron. The formula 

 (10) holds good for the exterior as well as for the interior of the 

 electron. The only difference between the two cases is this : The 

 limits T^, T.^, between which X does not vanish, are determined at a 

 point of the exterior by 



E^ — a = tTj R^ -\- a z= cr^ .... (11) 



7tj, 7?, meaning the distances of the point in question from the 

 position of the centre occupied at the time t- — t^ and t — t.^ ; on the 

 other hand for a point of the interior by 



