428 LIMITS OF OPTICAL CAPACITY OP THE MICROSCOPE. 



Instead of a series of lines printed letters may be used, the same 

 conditions being fulfilled, namely, by observing the point at sucb a 

 distance that the single letters may be just distinguished. On 

 looking at them through an aperture of 1 m.m. diameter, they will 

 be scarcely or not at all legible. This experiment is, however, not 

 so sensitive as the first. But, in all cases, the best accommodation 

 of the eye must be carefully maintained, otherwise the act of 

 passing a card, pierced with aa aperture, before the eye may, when 

 there is imperfect accommodation, actually improve vision by 

 diminishing the dispersion. 



The theory of diffraction of rays in the microscope loads, as will 

 be shewn in the following pages, to the conclusion, that any single 

 point of light in the object must, when viewed through the 

 microscope, appear exactly, as if an actual luminous point, situate 

 in the wia^e of the object, were observed through an aperture 

 corresponding in size and position to the ocular images (at 

 the so called eye spot) of the respective narrowest diaphragm 

 aperture. 



Hence it follows, firstly — that diffraction phenomena must be 

 visible when the ocular image has a diameter less than 1*89 mm., 

 and that the size of the dispersion circle, caused by diff'raction, 

 must increase in inverse proportion to the diameter of this ccul^r 

 aperture, consequently in direct proportion to the amplification, 

 supposing that the incident light from each point in the object 

 remains unchanged. Under such circumstances then, the image will 

 not, even with higher amplifications, suffer further loss of sharpness 

 of outline from diffraction, inasmuch as the dispersion circles 

 preserve, throughout, the same relation to the apparent magnitude 

 of the object. On the other hand, the deterioration arising from 

 diminished brightness and multiplication of darker entoptic shadows, 

 must increase with the amplification. From this it follows, 

 therefore, that, as a general rule, that amount of amplification 

 will shew most detail by which the minutest points that are visible 

 at all in the image, shall be presented under the most suitable 

 visual angle, namely, somewhat larger than that at which an 



