Diffraction Rings. By Alfred W. Porter. 5 



might be compared) which would have a perfectly definite mean- 

 ing, be totally independent of a particular observer's vision, and 

 at the same time represent the resolving power which a good 

 (though not phenomenally good) observer might be expected to 

 read. I have myself taught in my classes for ten years past that 

 this standard is purely conventional, and is easily surpassed. 



However, accepting Mr. Nelson's data, we must admit the 

 possibility of very considerably exceeding the conventional limit. 

 In order to meet such exceptional cases, I desire to propose a 

 new standard, which shall possess the merit of the old one of 

 being independent of the observer. Let the stars be brought to 

 such a closeness that the central depression just disappears ; it is 

 obvious that this closeness represents the " ne plus ultra " case of 

 resolution for monochromatic light. No one, however keen his 

 vision, will cpiiite succeed in seeing the star double at this limit- 

 ing distance. I propose, therefore, to take this degree of closeness 

 as the ultimate limit of resolving power. It corresponds to the 

 closeness for which the curves of intensity of the individual stars 

 cross each other at their points of inflexion (that is, at the points 

 at which they have no curvature). 



It is true that even for this degree of closeness, the oval shape 

 of the disk of light may enable one to infer that it is not a single 

 star which one observes. Moreover, if the light is polychromatic, 

 as usual, the tint at the centre of the resultant image may be 

 expected to be redder than on each side ; this, again, will tend to 

 make the limit of resolution lower than we would otherwise expect. 

 But the limit I here suggest is certainly so near the attainable 

 value, even when auxiliary circumstances such as these intervene, 

 that it is confidently put forward as the correct one to employ. 



