358 



Transactions of the Society. 



upper focal plane of the objective, where they interfere and remain in 

 the microscopic image, and, therefore, take part also in the final 

 virtual image. 



"If we shut out these interference images by diaphragms, or if 

 the angular aperture of the objective is too small to admit at least 

 the first pencil of rays produced by diffraction, as well as the undif- 

 i'racted light, the corresponding details disappear in the microscopic 

 image — a valve of Pleurosigma shows neither squares nor lines, and 

 fine rulings upon glass appear as a homogeneous surface." 



The development of this theory in full detail occurs at p. 232, 

 where they add the following diagram : — 



Fig. 68. 



In this diagram (fig. 68) A B is the optical axis of the Microscope ; 

 a the direct image of the source of light ; a' and a" are two diffraction 

 images of the same object the nearest to the direct image on either 

 side of it. B is a point in the plane of the image of the diffraction 

 grating which is taken to be the object on the stage of the Micro- 

 scope. The arc B P is drawn with a as its centre ; the other arc 

 which passes through the point P is drawn about a as its centre. 

 The length upon the axis A B intercepted between these two arcs is 

 taken to be one wave-length. 



1 do not propose to ask you now to wade through the mathematical 

 discussion of this diagram; it will suffice to. say that by geometrical 

 considerations Naegeli and Schwendener show that, upon certain 

 assumptions which are duly laid down, the distance from B to P 

 will be m times the distance of the bright lines in the object on 

 the stage which produces the diffraction, m being the coefficient of 

 magnifying power. It follows, as they say, that the number of 

 interference lines in any given space in the image plane agrees with 

 those of the strias. 



This is as far as their argument carries them ; that is to say, they 

 =diow that, in number and in distance from centre to centre, the 



