21 



AN OVERLOOKED POINT OONCERNING THE RE- 

 SOLVING POWER OF THE MICROSCOPE. 



By Julius Rheinberg, F.R.M.S. 



{Read December 18th, 1903.) 



The experiment I have the honour of showing you this evening 

 is a modification of one shown to me by Dr. G. Johnstone Stoney, 

 F.R.S., who, about nine years ago, made a most interesting 

 discovery, which, whilst fitting in perfectly with theory, seems 

 to have entirely escaped notice hitherto. It is this : — 



If we have a number of equidistant lines or points, it is well 

 known what the Numerical Aperture of the objective must be in 

 order to resolve them, and it has been tacitly assumed that, 

 whether the number of lines be two, three, four, or a large 

 number, so long as the distance between the individual lines 

 is the same, the same Numerical Aperture is needful for the 

 purpose of distinguishing the lines from one another. 



It has been left for Dr. Johnstone Stoney to demonstrate that 

 when there are only two lines, they can be resolved with an N.A. 

 sensibly less than that required to resolve a large number of lines 

 the same distance apart. 



The arrangement of the experiment on the table this evening, 

 showing this, is as follows. 



The object under the microscope is the 15,000 to the inch band 

 on one of those beautifully ruled test plates by Grayson, of Mel- 

 bourne. In this band, two adjacent lines happen to be somewhat 

 longer than the other ten, as seen in Fig. 1. An 8 mm. apo- 

 chromatic objective by Zeiss and a x 27 compensating eyepiece is 

 used. Just above the objective I have an arrangement like the jaws 

 of a spectroscope slit, which, actuated by the projecting wooden 

 tongs, can be opened and closed symmetrically from the centre of 

 the objective. The N.A. of the objective can thus be gradually 



