MICEOSCOPES: ALFRED W. PORTER 



69 



object. The condenser itself is an optical instrument to which the 

 principles of resolution apply. The greater the Numerical Aperture 

 (N.A.) of the condenser, the more nearly will each point of the 

 object be seen by light from a distinct point of the source; but 

 perfect independence is never secured. On the other hand, if no 

 condenser is used, or if it be not focussed for the object, each point 

 of the source will send light practically in one phase to a large 

 patch of the object. Other points will do the same. Thus the 

 independence between the lights at different points of the object 

 breaks down, and Abbe's result will be approximated to. That 

 is, for a dry objective (n = 1), instead of being able to resolve 

 lines separated by X/2 if sin a = 1, their distance apart will 

 require to be at least X. It is this halving of resolving power 

 which is brought about by replacing proper by random illumination. 



The difference between these two cases may to some appear 

 obscure. It depends on the fact that the light which passes through 

 neighbouring openings in the object spreads out by diffraction and 

 the diffracted beams overlap in the field of view. If there is no 

 definite phase relationship between these beams the case is analogous 

 to that of illumination by two candles — the intensities of light 

 can then be simply added together. When there is a phrase relation- 

 ship this is not lihe case. At points where there is an opposition 

 in phase the resultant amplitude may be zero. At intermediate points 

 the phase difference may be zero or a whole number of periods. In 

 this case the resultant amjiUtude is the sum of the separate ampli- 

 tudes and the intensity is the square of the amplitude. For the 

 sake of illustration two such superposed illuminations are shown 

 in the figure. The dotted curve represents the components placed 

 so that the maximum of one occurs at zero of the other. The 



