﻿Rectification of the Spectrum. 



951 



Measurements of N c were now made with the aid of a 

 Fraunhofer micrometer reading to a few ten-thousandths of 

 a centimetre. The opaque mirror was adjustably mounted on 

 a slide and moved normally to itself, in order to put the 

 centres of ellipses successively in coincidence with the 

 B, 0, D, kc. lines of the spectrum. At each coincidence a 

 reading was made, care being taken to proceed in one 

 direction only. As there are four spectra more or less in 

 coincidence, it is easily possible to select the wrong line. 



Figs. 6 & 7. 



mmj 



40 44- 4% 



By temporarily screening off the reflexion from one mirror, 

 however, and then marking the line desired of the remaining 

 pair by the cross-hairs of the telescope, mistakes are usually 

 avoidable. In the extreme blue and red it was often difficult 

 to see the lines clearly in the superimposed spectra, so that 

 the data for the B and G lines are not quite so trustworthy. 

 In fact the ellipses come out sharpest when but a narrow 

 strip of the grating is used ; whereas the resolving power of 

 the grating or general sharpness of lines increases with the 

 breadth of the grating. If the whole grating is used (an 

 intense spot of sunlight focussed on the slit), the adjustment 

 must be very perfect and the surfaces quite true to bring out 

 the cllipsss ; this is not the case if parallel rays fall on the 



