76 Dr. J. W. Draper on a new 



of its increasing intensity from the more to ihe less refrangible 

 regions is due to the compression of the coloured spaces that 

 correspondingly takes place, owing to the action of the prism 

 itselfj but having failed to obtain satisfactory measures in the 

 ease of the diffraction-spectrum (in which such compression or 

 condensation does not occur), I was led to reflect whether 

 better success might not be secured by attempting to measure 

 the relative intensity or distribution of light. 



Admitting what is commonly received as true, that the 

 yellow is the brightest of the coloured spectrum spaces, and 

 that the luminous intensity diminishes from that in both di- 

 rections above and below, I supposed that, if such a spectrum 

 were brought into the presence of an extraneous light, the 

 illuminating power of which could be varied at pleasure, after 

 the red and the orange on one side, and the green, blue, indigo, 

 and violet on the other, had been extinguished, the yellow 

 would still remain in the midst of the surrounding illumination. 

 On making the experiment it turned out differently. 



For the sake of clearness of description I will call this ex- 

 traneous light, from the functions it has to discharge, the ex- 

 tinguishing light. 



There are many different plans by which the principle above 

 indicated may be carried into practical effect. Several of 

 these I haye tried, and have found the following a convenient 

 one. 



Remove from the common three-tubed spectroscope its 

 scale-tube, and place against the aperture into which it was 

 screwed a piece of glass ground on both sides. In front of 

 this arrange an ordinary gas-light, attached to a flexible tube, 

 so that its distance from the ground glass may be varied at 

 pleasure. On looking through the telescope, the field of view 

 will be seen uniformly illuminated, this being the use of the 

 ground glass. The brilliancy of the field depends on the dis- 

 tance of the gas-light, according to the ordinary photometric 

 law. 



1. Case of the Prismatic or Dispersion Spectrum. 



If the extinguishing light be for the moment put out, and 

 in the proper place before the slit-tube the luminous flame of 

 the Bunsen burner that accompanies the apparatus be arranged, 

 on looking through the telescope a sjoectrum of that luminous 

 flame will of course be seen. The slit itself should be very 

 narrow, so that the spectrum may not be too bright. 



Now let the extinguishing flame be placed before the ground 

 glass, and a spectrum is seen in the midst of a field of light, 

 the brilliancy of which can be varied at pleasure. If this ex- 



