ON THE THEORY OF THE MICEOSCOPE. 207 



scope, or on the receiving surface of a camera ; and it was, there- 

 fore, tacitly premised that every optical function of the microscope 

 was determined, just as in these other instruments, hy the 

 geometrically traceable relations of the refracted rays of light. A 

 rigorous examination of the experiences upon which this tradi- 

 tional distinction of ''defining" and ''resolving" powers is 

 founded, has shewn that the proposition, though apparently con- 

 sistent with fact, is not admissible. It holds good, indeed, for 

 certain cases, capable of definite verification, but for the generality 

 of objects, and particularly for those objects on which the micros- 

 cope is supposed to exhibit its highest quality of performance, it 

 appears that the production of microscopic images is closely cm- 

 nected with a peculiar and hitherto neglected physical process, which 

 has its seatin and depends on the nature of the olject itself, although 

 the measure of its effect stands in direct dependence upon the 

 constraction of the objective. The results which follow from these 

 facts have a direct bearing on the most important problems in 

 micrography ; they shew the existence of an entirely specific 

 function of angular aperture, and in connection therewith present 

 clearer and truer notions respecting those two factors of so-called 

 defining and resolving power which constitute the optical 

 capacity of the microscope, and from the correct perception of 

 which every condition on which its performance depends may be 

 accurately determined. Hence, also, may be gathered practical 

 rules for the construction of the microscope on rational principles, 

 as well as suggestions for suitable methods of testing it when 

 made. On the other hand, an exploitation of the newly-gained 

 ground by farther experiments and researches led to deductions 

 respecting the general nature of microscopic vision. Thus it be- 

 came possible not only to fix the limits of the visible, beyond 

 which no farther resolution of microscopic structare coald be ex- 

 pected, but also to bring to light a fact of general application, 

 namely, that a microscopic image which may be entirely 

 free from error in itself, and therefore be supposed to represent in 

 all cases the true constitution of an object — a proposition on 



