On *' Central" Light in Resolution. By J. W. Stephenson. 39 



In fig. 3 in like manner the small circles represent at A' and B' 

 the still more efiective (because more oblique) portions of the illu- 

 minating beam, with their respective diffraction pencils a' and h'. 

 In this case the diffraction images within the limits of the back 

 lens are more complete, and either A' a or B' V would resolve the 

 striation on the valve, but a still more perfect image would be 

 obtained if by a suitable stop, one pair, say A' a, and all the useless 

 central light were shut out and the work done by B' h' alone, as in 

 fig. 4. 



Fig. 3. Fig. 4. 



With any given objective, on one and the same grating or 

 valve, the distance between the illuminating pencil and the dif- 

 fraction pencil thence arising is a constant quantity, and hence 

 it follows that to be effective, this distance must always be less than 

 the diameter of the back lens. Thus A a = B & = A' a' = B' &', 

 and just as A and B, or A' and B', recede from the centre on one 

 side, so are their respective diffraction pencils a and &, or a and h', 

 drawn towards it on the other, being as it were linked together by 

 this condition. Nor is this distance in any way affected by the 

 medium in which the object is mounted. In the cases which we 

 have discussed it is a matter of absolute indifference (if the object 

 adhere to the cover) whether it be in air, balsam, or phosphorus as 

 far as resolution is concerned, although its visibilitij, depending 

 as it does on the intensity of the lines, may be immensely 

 influenced thereby. 



It follows from what I have said that no objective with an 

 aperture less than 2*00 N.A, is capable of resolving an ordinary 

 valve of Amphipleura pellucida with a beam of " central " light. 



Very few persons possess objectives of 1 '50 N.A., but most of 

 the Fellows can verify the truth of the theoretical considerations 

 here put forth with an ordinary objective capable of resolving 

 Fleurosigma angulatum with oblique light. It will be found 

 that where a narrow central beam, from the smallest stop of a con- 

 denser, fails, the object will be immediately resolved by using a 

 beam of light of greater width, although from the flood of central 

 light the definition will be inferior to that obtained by a purely 

 oblique pencil. 



