497 



ON RESOLUTIONS OBTAINED WITH DARK-GROUND 

 ILLUMINATION AND THEIR RELATION TO THE 

 SPECTRUM THEORY. 



By W. B. Stokes. 



(^liead Jvne 2Wi, 1912.) 



Serious students of Microscopical Images have long sought for 

 a crucial experiment which shall decide the claims of the two 

 theories which have been put forward to explain them. Dis- 

 satisfaction with the theoretical conclusions of Abbe and his 

 supporters, and with their ideas with regard to the correct 

 management of the microscope, caused many to look to the 

 older theory propounded by Airy for an interpretation. Im- 

 partial authorities have, however, accepted Abbe's main results 

 for the reason that it is almost impossible to illuminate a 

 microscopic object without causing contiguous detail to transmit 

 light-waves which, having a common origin and consequently 

 a permanent phase-relation, are therefore in a condition to 

 interfere. A crucial experiment will therefore appear difficult 

 to make, and one would expect it to be made, if it is made, under 

 conditions which do not occur in ordinary practice. Leaving 

 aside the problem created by the use of a full or nearly full 

 cone of illumination, the solution of which has yet to come, 

 there yet remains one mode of microscopic vision which is 

 distinctly practical, and at the same time seems to offer the 

 very example which microscopists have sought. I refer to the 

 image obtained with dark-ground illumination. 



With this mode of illumination it is essential that the direct 

 light shall not enter the objective. If the object has a periodic 

 structure therefore, the so-called dioptric beam or spectrum of the 

 zero order will not be included in the objective unless the object 

 has some power of deflecting it by refraction or reflection. 



Leaving aside the possibility of such deflection, it would appear 

 necessary for resolution, if the Abbe spectrum theory is to be 

 applied, that a spectrum of the second order should be included 

 in the objective to co-operate with that of the first order. 



