Michelson — Spectroscope without Prisms or Gratings. 217 



The resolving power of such a combination is win, exactly 

 as in the case of gratings; so that with but twenty elements 

 5 mm thick and hence m = 5000 the resolving power would be 

 100000, which is about that of the best gratings. 



The experiment was actually tried with but seven elements, 

 placed between a collimator and an observing telescope and 

 the collimator slit illuminated with light from a sodium flame. 

 The images were so distinct that the broadening of the lines 

 could be very easily detected, and the Zeeman effect was 

 readily observed when the sodium flame was placed in a mag- 

 netic Held. 



It is important to note that the resolving power is inde- 

 pendent of the number of plates but depends only on the 

 total thickness, and the only advantage in a large number of 

 elements is the greater separation of the spectra. The 

 overlapping of spectra is doubtless a disadvantage, which 

 however could be overcome by a preliminary analysis ; and for 

 the examination of single lines and especially in the investiga- 

 tion of effects of broadening, shifting or doubling of lines, the 

 method seems especially well adapted. A battery of twenty 

 plates has been ordered and it is hoped that further results will 

 be obtained within a few weeks. 



Ryerson Physical Laboratory, 



University of Chicago. 



