Theories of Microscopical Vision. By A. E. Conrady. 625 



consist of a row of candle flames, if he had not failed to grasp 

 these elementary facts ! 



There is, however, an opening for a move reasonable objection 

 in this same direction, i.e. as to the rendering of number, length 

 and possible irregularities of the lines in the image. We may very 

 well ask : — 



1. How is it that the waves which come into interference in the 

 plane of the image do not extend their effect beyond the geometrical 

 image ? Under my method of treating the diffraction- images the 

 answer seems almost obvious, for it shows that there is maximum 

 brightness only along those lines of the image which are conjugate 

 to the centre lines of the slits of the grating ; hence, there can 

 be only the correct number of fully bright lines, and they must be 

 shown at very nearly their exact length. According to Abbe's 

 treatment, where the waves leaving the object are followed as 

 such through the optical system, the answer follows from the 

 Huyghenian principle which shows that waves in their progress 

 conform closely to the outline of the corresponding " rays " of 

 light, with but slight encroachment beyond that outline ; hence, 

 on reaching the plane of the final image, the waves will, with the 

 exception of a small fringe, again be confined within the limits of 

 the geometrical image of the object. In either case the formation 

 of one or possibly several feeble ghost-lines beyond the limits of 

 the geometrical image appears probable ; and I believe that, as a 

 matter of fact, feeble spurious lines of this kind are occasionally 

 observed, even under ordinary working conditions. 



2. As to the rendering of irregularities in the lines, the ex- 

 planation is contained in one word : " Ghosts." I mean, of course, 

 the " ghosts " painfully familiar to users of diffraction gratings for 

 spectroscopical purposes. The sweet simplicity of the diffraction 

 phenomena which results from the assumption of an absolutely 

 regular and uniform grating, is disturbed by any irregularity in 

 the ruling. The light from any displaced slit, or from a slit differ- 

 ing in width from the majority, does not completely harmonise 

 with that from the other slits, and there results a corresponding 

 zone of greater or less brightness in the diffracted waves, besides 

 uncompensated, scattered light from the irregular slit ; all these 

 results combined, form in the first place so-called "ghost" spectra 

 in the principal focal plane of the objective, and subsequently 

 lead to a more or less faithful representation of the irregularity 

 in the final images. 



Leaving some other objections to be dealt with later, I come to 

 a very serious one which, curiously enough, is the very one which 

 has escaped Mr. Gordon's attention, and which, as far as I know, 

 has formed the only really formidable barrier to reconciling a 

 direct and necessary conclusion from Abbe's theory with the 

 experience gained in ordinary microscopical observations. I refer 



