328 INFRA-RED EMISSION SPECTRA. 



pressure, on account of the scarcity of the molecules, there will be fewer 

 collisions in a given time, the ionization will decrease, and the "electri- 

 cal temperature" will decrease. 



On the other hand, this explanation will not account for the behavior 

 of the 4.75 IX band, which appears to be due to a thermal radiation, 

 excited by the collision of electrons with the neutral gas molecules. 

 The gas molecule as a whole will suffer an increase in its kinetic energy, 

 and, in colliding with other molecules, will cause a rise in the thermal 

 temperature of the gas. With increase in pressure, i. e., in the number 

 of gas molecules, the number of collisions will increase and the intensity 

 of the thermal radiation will increase, but will not pass through a maxi- 

 m.um, as is true of the other bands, because a stage will be arrived at 

 where there will be a decrease in the ionization and in the collisions of 

 the molecules. At still higher pressures the gas would cease to con- 

 duct the current. 



Aside from these theoretical considerations, there is some experi- 

 mental evidence for believing that the 4.75 jx band is of thermal origin. 

 First, the gas must be hotter than the tube, for during the passage of 

 the current the radiation tangential to the axis of the tube is probably 

 different from the longitudinal, and the cell-walls assume the mean 

 temperature of the gas only after the current has passed for some time. 

 It has already been shown under the discussion of the temperature of 

 the gas in the vacuum-tube (p. 322) that the mean thermal tempera- 

 ture is from 300° to 400° C, depending upon the current and the press- 

 ure. It was also shown there that the black body at these temperatures 

 did not emit a perceptible radiation at i fx, while the emission lines in 

 this region are very intense, indicating a temperature of perhaps 4,000" 

 abs. On the other hand, the "black body" at 4.75 /a radiated almost as 

 intensely as the vacuum-tube. This would indicate two distinct tem- 

 peratures, which is hardly the case, and the term "electrical tempera- 

 ture" seems appropriate for the intense lines at i /x. 



Second, the distribution of the heat in the vacuum-tube is very dif- 

 ferent for different pressures. At a high pressure the constricted por- 

 tion of the vacuum-tube is the hotter, while at low pressure it is quite 

 cool, and the region surrounding the electrodes is the hotter. 



Third, the emission increases with the pressure (equivalent to an 

 increase in the thickness of the emitting layer) and approaches a limit- 

 ing value. Paschen (loc. cit.) has found for CO2 at atmospheric 

 pressure, in a brass tube heated by a Bunsen burner, that a column 7 

 cm. long emitted and absorbed energy just as strong as a column 33 

 cm. long. In the present case the 15 cm. column of CO, at 5 mm. 

 pressure is equivalent to a column i mm. long at 760 mm. pressure. 



Stark, Elektricitat in Gasen, from consideration of the kinetic energy of the elec- 

 tron computes an electrical temperature of some 6,000°. 



