2^22 INFRA-RED EMISSION SPECTRA. 



The fact must not be overlooked that the radiation from the vacuum- 

 tube gave large deflections simply because of the great sensitiveness of 

 the instrument and not on account of the actual intensity as compared 

 v^^ith the deflections from gases in a Bunsen flame. In the latter the 

 deflections for the 4.4 jU. CO, band were some 60 to 70 cm. for the old 

 instrument, so that in the present work the deflections would be 50 

 times as great, viz, 3,000 cm. In other words, in the vacuum-tube 

 curves the deflections are only from ^l^ to 30'bo as great as from the 

 Bunsen flame and the Nernst heater. 



temperature; of gas in the vacuum-tude. 



Warburg's theoretical work on the temperature of the vacuum-tube 

 has already been mentioned. He showed that the temperature of nitro- 

 gen is much higher than that of hydrogen. 



The intensity of the 4.75 /x impurity band in nitrogen, for all pres- 

 sures, was so much greater than for all the other gases that one is led 

 to think that it is due to the higher temperature of the gas. 



Wood (loc. cit.) gives data for the observed mean temperature of 

 gases in a vacuum-tube for different pressures, and currents of 0.001 

 to 0.003 ampere. In all cases the computed and observed temperatures 

 are in fair agreement, the observed values being slightly less than the 

 computed values, as one would expect, from the use of a thermopile 

 which can not be made infinitely thin. The observed values fall upon a 

 straight line, which shows the accuracy of the observations. Using 

 these values and extrapolating to a current of 0.02 ampere, in the 

 present work, for a pressure of 1.8 mm. of nitrogen this would indicate 

 a temperature of about 250° C, while for a pressure of 3 mm. the tem- 

 perature of the axis would be about 325° C. For a current of 0.025 

 ampere the temperature would be 300° and 400° respectively. 



The whole shows that for the large currents used (0.02 to 0.028 

 ampere) it is not unthinkable that the 4.75 \x band is due to the heating 

 of the residual gas. On the other hand, the black body at 325° does 

 not emit a very perceptible radiation at i [i, while the emission lines in 

 this region are very intense. They indicate a very much higher tem- 

 perature — some 4,000° abs.,if we consider the maximum of the envelope 

 of the curve, drawn through the highest points of these emission lines, 

 which maximum lies just beyond the red. If the gas had this high tem- 

 perature then one would expect the residual CO molecules to grow 

 hotter. This reasoning leads to the result that the gas is at two distinct 

 temperatures, which is hardly the case. Electrical excitation suggests 

 itself, which brings us to a theoretical consideration of the phenomenon 

 of vacuum-tube radiation. 



