RADIATION FROM A VACUUM-TUBE. 



319 



(fig*. 151) some of the latter vapor must have been present. The second 

 sample, made with greater precautions, did not show any radiation 

 beyond 2.5 /x. In the region of 4.75 /*, curve b, the observations for 



<ti 



the radiation from the hot cell and the gas (indicated by circles and 

 crosses) coincide. The nitrogen bands at 0.75 ju-, o.g fi, and 1.06 /a are 

 due to the dissociated NHo. 



RADIATION FROM A VACUUM-TUBE WHEN HEATED EXTERNALLY. 



The very different behavior of the 4.75 fi band from those at the end 

 of the red made it highly desirable to learn whether it can be due to the 

 mere rise in temperature of the gas. If it obeys Kirchoff's law, it can 

 not be more intense than the black-body radiation at the same tempera- 

 ture. The radiation of the hot glass cell beyond 3 /x, when hot (at 

 cathode rays pressure) gave deflections that were comparable with that 

 of the 4.75 IX band. This would lead one to think that the 4.75 [x band 

 is due simply to rise in temperature of the gas. On the other hand, in 

 the region oi i /x the hot cell gave no appreciable deflections, while the 

 gas (nitrogen) gave strong emission bands, which would indicate that if 

 these emission bands be due to a pure thermal excitation, following Kir- 

 choff's law, then the temperature of the gas must be very high as com- 

 pared with that of the cell walls. The problem then was to find the 

 radiation from the vacuum-tube when heated externally to the tempera- 

 ture it had during the electrical discharge, which is really the black- 

 body radiation ; also to find the radiation from an approximate black 

 body, e. g., a Leslie cube, the temperature of which can be accurately 

 determined. 



In order to determine the radiation from the vacuum-tube when 

 externally heated, the part between the electrodes was wound with an 



