q: 



:= 



()-25 



Q'l- 



— 



18-7' 



Qo" 



^ 



232-0 



<. 



- 



174-0 



Q-. 



= 



304-0 



q: 



— 



U56-0 



t. 



^ 



893-0 



Q-. 



^ 



4100-0 



Aberration of the Concave Grating. 135 



For the extreme field S*''' the ratios run 



(1) A-Btype 



(2) C-D type 



(3) E type 



The more favourable showing for the O.S. types of 

 mounting for large angles of diffraction is due, in the case of 

 the A-B type^ to a real improvement in definition already 

 noted ; in the other case the improvement is only apparent 

 and comparative, and not real. For type C-D the total 

 aberrations are about ten times as large for fi.eld S^^ as for 

 field S^ and for type E, about thirty times as large for S^'^ as 

 for S'. ' 



These comparisons, particularly those for S' and S^^, the 

 fields most frequently used with the O.S. types_, show very 

 clearly the great inferiority of the concave-grating objective 

 spectroscope, as an optical instrument, to the ordinary form 

 of concave-grating spectroscope, and are a sufficiently complete 

 explanation of the disappointing results that have been 

 secured by its use in solar eclipse w^ork. 



Poor and Mitchell considered the star-spectrum plates 

 which they obtained (using the A-B and the F types of 

 mounting) satisfactory, but it must be remembered that such 

 plates afford no such refined tests of definition as plates of 

 solar spectra. Their results are considered more at length 

 later. 



Although we are thus forced to the general conclusion that 

 the concave objective grating spectroscope is unsuited to the 

 class of work that can be performed by the same gratings 

 when used on Rowland mountings, the advantages of the 

 former type of instruments are so great in certain cases that 

 it becomes desirable to investigate the question as to whether 

 there is not some way of reducing the aberration in these 

 instruments. 



