CHAPTER II. 



SCOPE OF PRESENT INVESTIGATION. 



The present work divides itself into two parts: (i) emission spectra 

 of the arc between metalHc electrodes and the emission spectra of the 

 chloride of the metals in the carbon arc; (2) emission spectra of gases 

 in vacuum-tubes. The contrast between these two forms of radiators is 

 worthy of notice. The arc is noted for its enormous heat radiation in 

 proportion to its light radiation. On the other hand, the vacuum-tube 

 radiates but little heat. Consequently, in the study of these two kinds 

 of radiators, the form of the device for exploring their spectra must 

 differ. If a radiometer is used its period must be short for examining 

 the arc in order to avoid heating of the window and the consequent shift- 

 ing of the zero reading. This, however, is of less importance than the 

 variation in intensity of the radiation from the arc, which requires a 

 recording instrument having a short period. For the vacuum-tube a 

 much greater sensitiveness must be used, which means a longer period 

 for a linear radiometer vane. Fortunately the radiation from the 

 vacuum-tube is uniform, which permits the use of a slow-period instru- 

 ment. 



THE radiome;te;r. 



In the present investigation a Nichols' radiometer, a 7 cm. rock-salt 

 prism, and a 35 cm. focal length mirror spectrometer were used.^ 



As a device for exploring infra-red spectra the radiometer has some 

 inherent advantages as well as disadvantages over the bolometer. The 

 most prominent advantages are its freedom from magnetic disturbances 

 and its ease of construction and use at a high degree of sensitiveness. 

 Its lack of mobility is not objectionable in the present work. It is also 

 of interest to notice that in all my infra-red work this radiometer has 

 stood upon a table, which itself stood upon a cement floor of a base- 

 ment room, without any protection against earth tremors. Neverthe- 

 less, only on rare occasions were the observations interrupted by earth 



^E. F. Nichols: Astrophys. Jour., 13, loi, 1901. 



^Coblentz : Phys. Rev., 16, p. 35, 77, 1903 ; also in "Absorption Spectra." (Wash- 

 ington, 1905). 



295 



