150 



VII. — A Correction for a Spectroscope. 

 By Edward M. Nelson. 



(Read December 18, 1907.) 



It is the common experience of everyone who has worked with a 

 spectroscope that the image of the slit is represented by curved 

 lines, especially when high up in the spectrum. Now curved 

 images are due to the spherical aberration of an oblique pencil. 

 Therefore we know by the curved image upon the plate that we 

 are dealing with an oblique pencil ; and although rays which have 

 been parallelized by the collimator and passed through the prism 

 are supposed to fall upon the telescope in a direct manner, a little 

 consideration will show that the prism has, by its refraction, never- 

 theless rendered their incidence oblique. The correction for this 

 «rror is obvious. The telescope objective should be mounted upon 



Fig. 30. 



a pivot so as to be capable of rotation about a vertical axis ; an 

 indicator pointing to an arc, graduated in wave-lengths, would be 

 convenient for setting the objective at any required position. It 

 would be desirable and very simple to make this adjustment 

 automatic (see fig. 30), by fixing an arm B to the pivot A, carrying 

 the object-glass of the telescope, and by means of a spring making 

 this arm bear upon a horizontal excentric, C, fixed to the axis of 

 the pillar D. 



Then, as the telescope was rotated round the axis D of the 

 pillar, the arm B would be moved by the excentric C, and the 

 object-glass turned upon its pivot A. 



By this means, the lines in a spectrum would be rendered 

 perfectly straight, because the incidence would be always direct, 

 and what is more important, the lines would be made critically 

 sharp. 



It is difficult to understand why spectroscopists have for so 

 long been content with a curved image of a straight object, and 

 fuzzy images. 



