spherical Wave of Light. 6;^ 



of surface only, we integrate over a finite portion we get a 

 corresponding result. Hence it would seem probable either 

 that n is greater than 2 or that h = 0. 



Returning to the consideration of the problem, imaginino- 

 disturbances of the form which I have found to originate 

 from every portion of the spherical surface, and finding their 

 resultant at some point, there appear in the expressions, 

 (i) terms depending upon A/<: where c is the radius of the 

 original sphere, which terms can not be made to disappear, 

 and which express the expected fact that we can only apply 

 the principle to a wave of which the radius is large compared 

 with the wave length, and (2) terms depending upon X/r 

 where r is the distance of the point considered from the sur- 

 face of the sphere, which all disappear upon integration by 

 a continual appearance of a set of similar series. As these 

 terms disappear we obtain from this investigation no criterion 

 of the value of n or of d, but the fact of their disappearance 

 gives a strong confirmation (which appears in other ways 

 in the paper) of the appropriate nature of the solution of the 

 differential equation used, and of the possibility of satisfying 

 the conditions without taking into consideration any wave 

 of condensation. The only conditions which we thus 

 obtain are 



i;(-i)X = 0. 

 I 



A 



In the last part of the paper I consider the case of a 

 wave converging to the origin. 



The wave being the secondary wave representing a plane 

 polarised wave, reversed by the reversal of the velocity, I show 

 that the resultant displacement and rotation at the origin 

 will be of the same nature as the original displacement at 

 the origin, if n have the values 2 and 3, and ^2 = 0. 



^2 = ^3 = ^3 = -.- 

 2 A. 



