142 



MK. CHARLES DE WATTEVILLK ON FLAMK SPECTRA. 



solutiun shall reach the flame. The sprayer described by M. GOUY works very well 

 in this respect ; that is to say, a relatively small quantity of air is able to produce a 

 sufficiently large amount of spray. In order to ensure this condition, the annular 

 space between the end of the inner tube and the inner wall of the outer tube should 

 be made as small as possible. 



II. SPECTROSCOPICAL APPARATUS. 



Both diffraction grating and prism spectroscopes have been used in this research. 

 The grating offers a ready means for the direct determination of wave-lengths, and, 

 moreover, on account of its high resolving power, enables closely adjacent lines to be 

 easily separated. This is important, since whenever a particular line is built up of several 



Fig. -2. 



components, it is necessary to determine whicli of these components are present in the 

 flame. On the other hand, with the grating, a large amount of the incident light is 

 uselessly spent in the formation of the spectra of different orders. Consequently the 

 prism spectroscope has been employed for mapping these lines which are too weak to 

 be detected on the photographs taken with the grating spectroscope. 



