1897-98-] Microscopy and some of its Uses. 315 



In physical optics it is shown that when light passes through 

 a very tine grating, or very fine lines ruled on glass, it does 

 not, after emerging from the grating, continue to be propagated 

 in straight lines. It spreads out sideways all round, and we 

 have to think of it as radiating from every point on the emer- 

 gent side of the grating in an infinitude of wide conical 

 pencils. This diatom, then, with its fine symmetrical lines, 

 like nearly all the objects looked at by microscopes, is com- 

 parable to a very fine grating, and light behaves similarly in 

 passing through it. In the innumerable rays which radiate 

 from this diatom we may, however, distinguish two kinds. 

 There are first those which are but little bent from a straight 

 path, — these go to form the outlines of things, and can be 

 grasped by lenses of small angle. Then there are those which 

 are very much bent or diffracted in all directions, — so much so 

 that only part of them can be grasped by the widest-angled 

 objectives, — and the experiments of Abbe show that it is these 

 diffracted rays which chiefly image fine structures and details, 

 and that unless an objective has sufficient width of angle to 

 grasp a considerable portion of them it cannot have high 

 resolving or defining power. 



Like most discoveries, this is only a specific application of 

 a general natural principle. In ordinary vision, if we were to 

 think away all the infinitely small and irregular particles 

 which make up a surface or a body, we should see nothing. 

 It is because bodies are never perfectly smooth or perfectly 

 uniform that they are visible. In fact, it is the infinitely 

 small irregularities of a surface which, acting like prisms and 

 lenses, reflecting, refracting, and diffracting light, form images, 

 and constitute the details of all the forms we see. There is 

 thus no escape from Abbe's theory, and the process of super- 

 position of images is just like that of an artist first sketching 

 the outlines and then filling in the details. Hence a micro- 

 scopic image can only be true to nature when all the diffracted 

 rays are included : until this can be accomplished we are 

 only " seeing through a glass darkly " — the images are true 

 so far as they go, but not the whole truth. 



But though the general theory of diffraction is of very great 

 importance, and while I have wished to accentuate the advan- 

 tages of kindred scientific knowledge, I hope I have not left 



