On the Effect of Illumination, etc. Bi/ Prof. E. Abbe. 723 



beams of which a wide incident cone is composed, are dissimilar 

 images in general ; and this for two reasons: — 



(i.) Provided the diffraction effect of the structure is not confined 

 to very sm-ill angles, the admission to the objective of pencils of 

 different ohliqiiitij (i.e. pencils which result from incident rays of 

 ditiierent obliquity) is different — or a different part of every elementary 

 diffraction-fan is lof-t. That belonging to an axial incident beam is 

 admitted in its central position, and a peripheral portion only is lost ; 

 that which belongs to an oblique ray is stopped off up to a full half, 

 one from the left hand, another from the right. As the admission of 

 different parts of the total diffraction pencils of a structure makes the 

 images always different (at least in the more minute features), their 

 different obliquity of incidence alone would be sufficient to make the 

 resulting image a mixture of different (and, in general, differently 

 dissimilar) images ; and this will be so, even in the cases where the 

 diffraction-effect itself (not considered the different mode of partial 

 admission to the aperture) is the same for an axial and an oblique 

 incident ray. 



(ii.) But this diffraction effect by itself is not the same for rays 

 of different obliquity, except with such structures as act solely by 

 means of absorption of light (total or elective), i. e. with such struc- 

 tures the detail of which shows simply a difference of transparency 

 of the elements. In all structures, the elements of which show at the 

 same time difference of refraction (or of density), rays of different 

 oliliquity are subjected to different amounts of retardation during the 

 transmission ; and owing to this, the diffraction-pencils arising from 

 incident rays of different obliquity are of unequal constitution. Theory 

 and experiment show that the images, projected by meaus of two 

 beams of different obliquity of a structure of that kind, may be very 

 widely dissimilar, in such a degree that in one partial (elementary) 

 image there is a maximum of light at the same point of the field 

 where the other has a minimum, or darkness. The mixture of a 

 variety of elementary images of that kind must, consequently, produce 

 more or less confusion, or even total suppression of structural detail. 



This is practically made use of with great success in the method 

 of R. Koch of Berlin, in the observation of tinted preparations. He 

 recommended, twelve years ago, the use of a wide-angled cone of ligbt, 

 in order to suppress in the image of a preparation all the elements 

 which have no colour, and to enhance in that manner the image of 

 the coloured elements. The latter ones act merely by absorption, 

 and therefore give rise to equal diffraction spectra for rays of different 

 obliquities ; the former ones (the histological tissues, uncoloured) act 

 merely by means of different refraction and different retardation of the 

 transmitted light, and give rise to dissimilar diffraction-pencils and 

 dissimilar elementary images, the mixture of which is obliteration. 



(5) The result of this consideration is : — The resulting image, 

 produced by means of a broad illuminating beam, is always a mixture 

 of a multitude of partial images which are more or less different (and 



3 E 2 



