482 PROCEEDINGS OF THE SOCIETY. 



the sharply defined spectral images of the light source. But we can 

 cause a fresh set of sharply defined visual spectra to appear under these 

 conditions by contracting the iris diaphragm or stop under the con- 

 denser, and these will be formed in the region of the upper focal plane 

 of the objective. 



I must not stop now to discuss further how this can be brought into 

 harmony with the principles of the Abbe theory ; the only point I wish 

 to draw attention to for the moment, is that we cannot apply the method 

 of visual demonstration devised by Prof. Abbe, if we set up entirely 

 different conditions ; and this no doubt accounts for Dr. Zimmermann 

 and other followers of the Abbe school not having set up their experi- 

 ments that way. 



I will now proceed shortly to discuss one of Mr. Gordon's own very 

 instructive and interesting experiments. 



He shows that, under suitable conditions, the image of a Zeiss diffrac- 

 tion plate and that of an aerial image projected by the condenser on 

 the object plane of the Microscope, will behave in precisely the same 

 fashion. This is an experiment with which I am familiar, having 

 myself made a great number in similar fashion. This experiment 

 certainly does prove that, given suitable diaphragms above the objective, 

 the same effects can be produced with self-luminous objects as with 

 gratings. It proves that, so far as self-luminous objects are concerned, 

 the diaphragm above the objective must produce false effects. But so 

 far as the grating is concerned, it only proves that the diaphragms may 

 produce false effects. The grating, according to the way in which it is 

 illuminated, may cause the light to be unevenly distributed in the upper 

 focal plane of the objective, and if the diaphragm happens to be so 

 placed that its openings just coincide with the positions of the maxima, 

 the presence or absence of the diaphragm makes no difference, for it has 

 not then stopped out any light. 



With the self-luminous source the whole upper focal plane of the 

 objective is necessarily filled with light, and the diaphragm must there- 

 fore necessarily stop out some of it. The remainder which is passed 

 through may, under suitable conditions, correspond to the whole of that 

 which was emitted from the grating. Therefore the effect would be the 

 same. In fact, it is like using a stencil which equalises the pattern of 

 light in both cases, though the opaque parts of the stencil may cover 

 things in the one case which they do not in the other. 



Of course, so soon as parts of the co-operating spectra * emitted by 

 the grating are stopped out, then the diaphragm is responsible for the 

 altered image effect, and this accounts for the duplication, triplication, 

 serrated edge, and all the other effects. But that is in accordance, not 

 in discordance, with the Abbe theory. 



1 do not propose to discuss in detail the remainder of Mr. Gordon's 

 paper. With much of it, and particularly his admirable constructive 

 treatment of the image as a collection of antipoints, I thoroughly agree, 

 only I think this theory must be widened out by admitting the factor of 

 the diffraction by the object as well as that produced by the diaphragm, 

 and I think I have shown that the arguments do not touch this important 

 fundamental principle of the Abbe theory. 



* By co-operating spectra, such spectra are meant as have originated from one 

 point of the light source. 



