74 



VISION WITH THE COMPOUND MICROSCOPE 



around the direction of the incident beam from which they originate. 

 In such a case even small apertures are capable of admitting the 

 whole. The images of such coarse objects will therefore be always 

 perfectly similar to the object, and the result of the interference 

 effect is the same as if there were no diffraction at all, and the 

 object were a self-luminous one. 



(4) ' When the elements of a structure are reduced in diameter to 

 smaller and smaller multiples of the wave-length which corresponds 

 to the medium in which the object is, the diffraction pencil originating 

 from an incident beam has a wider and wider angular expansion 

 (the diffracted rays are further apart) ; and when they are reduced 



to only a few wave-lengths, not even 

 the hemisphere can embrace the whole 

 diffraction effect which appertains to 

 the structure. In this case the whole 

 can only be obtained by shortening 

 the wave-length, i.e. by increasing 

 the refractive index of the surround- 

 ing medium to such a degree that the 

 linear dimensions of the elements of 

 the object become a large multiple of 

 the reduced wave-length. With very 

 minute structures, the diffraction fan 

 which can be admitted in air, and 

 even in water or balsam, is only a 

 greater or less central portion of the 

 whole possible diffraction fen corre- 

 sponding to those structures, and which 

 could be obtained if they were in a 

 medium of much shorter wave-length. 

 Under these circumstances no objec- 

 tive, however wide may be its aperture, 

 can yield a, complete or strictly similar 

 image! 



It is at points of such extreme 

 delicacy and moment as this that the 

 diffraction hypothesis of Dr. Abbe is 

 so liable to misapprehension and mis- 

 interpretation, and a further note from 

 him relating to the dissimilarity of the 

 image in the case of incomplete admis- 

 sion of the diffraction pencil will be of 

 great value here. 



i. 'In the case of regular periodic 

 structures (i.e. equidistant striae, rows of apertures, ' dots,' and so forth) 

 the distance of the lines apart is, even with an incomplete admission 

 of the diffracted light, always depicted correctly ; that is to say, the 

 number of the lines per inch is never changed, provided the direct beam 

 (i.e. the central maximum of the diffraction fan) is admitted to the 

 objective and at least one of the next diffracted rays,, or, in^other words, 

 one of the next maxima of second order. The range of dissimilarity 



