Form of Spectrometer. 79 



the condition of their sight, the extinction took place in the 

 same manner. I had not an opportunity of examination in a 

 case of colour-blindness. 



2. Case of the Grating or Diffraction-spectrum. 



If the cause of the increasing intensity of light, in the 

 prismatic spectrum, from the more to the less refrangible 

 region, be the compression exercised by the prism on the 

 coloured spaces, increasing as the refrangibility is less, we 

 ought not to find any such peculiarity in the diffraction- 

 spectrum. In this the coloured spaces are arranged uniformly 

 and equally in the order of their wave-lengths. An extin- 

 guishing light ought to obliterate them all at the same 

 moment. 



Having modified the common three-tube spectroscope, as 

 has been described, I put in the place of its prism a glass 

 grating inclined at 45° to rays coining in through the slit. 

 The ruled side of the grating was presented to the slit. Now, 

 when the extinguishing flame was properly placed before the 

 ground glass, the plain side of the grating reflected its light 

 down the telescope-tube. In this, as in the former case, the 

 spectrum was seen in the midst of a field of light, the intensity 

 of which could be varied by varying the distance of the ex- 

 tinguishing flame, or by varying the opening of its stopcock. 

 This light needs no reinforcement by increasing the reflecting 

 power of the back face of the grating, these spectra being 

 much more feeble than those given by a prism, and the un- 

 assisted light being quite able to extinguish them. 



As the glass grating I was using gave its two series of 

 spectra of unequal brightness, I selected the most brilliant, 

 and in it used the spectrum of the first order. I saw, not 

 without pleasure, that as the force of the extinguishing illu- 

 mination increased, all the coloured spaces yielded apparently 

 to an equal degree, and disappeared at the same moment. 

 Sometimes, however, there seemed to be a very slight dif- 

 ference in favour of the red. On diminishing the illumination 

 all the colours came into view, apparently at the same time. 

 This spectrum gives a better opportunity than the prismatic 

 for observations on the yellow space, which, by being uncom- 

 pressed, exposes a wider surface to view. This yellow space 

 showed no superiority in resisting extinction over the other 

 colours. 



But as gaslight compared with sunlight is deficient in the 

 more refrangible rays, I repeated the examination of the 

 latter, as I had previously done for the prismatic spectrum, 



