TRA>SAC'riONS Ol'" SECTIOX A. 613 



no srcat aa a liuiiJred millious per second, thic becomes 4.10 *. When the wires 

 are close together, an extra term R must be added to L. Tor t-wo \\ire3 eijual in all 

 respects, R is the real part of 



and till' arniiment of the P.osscl fimctiony is in all cases 



'•■'""===VT-(v'')- 



This neglects a term of order (') , and may be used practically with suf- 



licient accuracy even when c = iia. 



If the wires are Aery far apart, so that, although 7rirji" is negligible, n-c'-lu" 

 is not, then, subject to an error of order jva-ji" relative to unity, 



L^oy /""cx _4/io„ «£ , .577 1 , V b err ber'. r + bei.r bei\r 

 ~ ^ \v / I " iiv J X (ber' ,r)- + (bei' .r)- 



where y^ is the ordinary Bessel function of zero order of the second hind, 

 defined by 



I o>i on J »=-o 



whose tabulation is well known. 



3. On the Ether Stress of Gravitation. i»y Professor F. Purser, F.T.C.D. 



The problem of which Maxwell first attempted the solution, of accounting 

 for gravitation by a stress of the ether, is that of solving the equations 



(1) 



p being the density of gra\itating niattei', V tiie gravitation potential, 

 A, B, C, F, G, H, according to Dr, Williamson's notation, the six components 



of stress. Substituting for p - _. y-V, it is easily seen that these become 



47r 



dA^dll^dX^^dAJ ^dW ^dG^ 

 dx dy dz dx dy dz 



dR dB d¥ dR' dB' , dY' /.,v 



dx dy dz dx dy dz 



dQj dV dC dG' dF' , dC 



— + — ^ = — + — + — . 



dx dy dz dx dy dz 



