^o6 



INFRA-RED EMISSION SPECTRA. 



marked contrast with the radiation from the electrodes, any trace of 

 which, as already mentioned, was sufficient to cause large deflections. 

 For the black body the radiation is not very intense at a low tempera- 

 ture and the maximum lies beyond 3 /x. The oxides in the arc have a 

 high temperature ; the maximum lies at 2 /x, and the weak radiation in 

 this region is to be found attributed to the low density of the vapors. 



Sodium (Na). 



Snow (loc, cit) found that the salts of the metals gave the same 

 emission lines as the metal itself, and that the chlorides were well 

 adapted for emission-spectra work, using hollow carbon electrodes. 

 For this reason only the chlorides of the metals were used. The present 



MM 



16 



2 



f • j [ 1 ■ — '- 



3 4 



Fig. 138. 



7^ 



work was not concerned with the verification of Snow's results, which 

 were used only in comparing the relative intensity of the lines investi- 

 gated. He used a quartz prism, which on account of its larger disper- 

 sion permits a greater accuracy in determining wave-lengths. In the 

 present work the question was whether there are lines beyond the region 

 investigated by him. The distribution of the energy in the vapors of 

 Na is shown in figure 138, where the ordinates are deflections in milli- 

 meters. 



Snow found the intensities of the lines at 0.589 ,«,, 0.818 /x, and 1.13 ^i 

 to have a ratio of 87 : 66 : 42. In other words, the energy in 0.589 /* is 

 more than twice that of the line at 1.132JU,. In the present work the 

 intensities of these two lines are exactly reversed, and in the same pro- 

 portion. Since the dispersion is smaller this may be due to the impurity 

 of the spectrum. 



