DELINEATION OF THE FINE STRUCTURE OF OBJECTS. 235 



correspond to half a wave-length, with illumination which we 

 assumed to be as oblique as possible. 



Wave-length in Mic. Limit of Discrimination in Mic. 



Red '68 -34 



Yellow -58 -29 



Green '52 '26 



Blue '43 '22 



Violet -39 -20 



Line M '37 '19 



Line N *36 '18 



The figures given for the lines M and N are taken from the paper 

 of Draper. 1 The so-called chemical rays therefore approximately 

 coincide with the most refrangible rays. In photography, con- 

 sequently, the conditions are about the same as would be given by 

 direct vision through the Microscope by a structure coarser in the 

 proportion of 2 to 3. 



The proof that the delineation of fine structures is not brought 

 about dioptrically, but through the interference of diffracted rays, 

 is of great import as regards the interpretation of the microscopic 

 image. For while a dioptric image, as produced by homofocal 

 pencils of rays, is point for point similar to the object situated in 

 the plane of adjustment, and admits therefore of a wholly reliable 

 conclusion as to the composition of the object, if a correct stereo- 

 metric interpretation is made of what is seen on the surface, the 

 interference images produced by the process of diffraction have no 

 constant relation to the nature of the corresponding object. For 

 instance, rows of points produce the same image as actual lines, and 

 where two such series of lines intersect at right angles two further 

 lines appear in the interference image in the direction of the 

 diagonals. Lines which lie in the same level are often rendered 

 visible or invisible by varying the adjustment. It may occur 

 that a simple series of lines is delineated as of doubled or trebled 

 fineness. All these phenomena may at once be placed upon a 

 theoretical basis ; they belong to the known class of diffraction 

 phenomena produced by gratings, which are treated of in every 

 text-book of Physics. This implies that the structural indications 

 of the composition of the object, as formed in the microscopic 

 image, occasionally conform to the original; as a rule, however, 

 they do not so conform. It is clear, under these circumstances, that 



1 Poggendorff's " Annalen," Bd. cli. (1874), p. 337. 



