GASES IN VACUUM-TUBES, 



311 



it is not a great amount in comparison with the total deflection. Since 

 the cell became heated the least at 0.6 to i mm., it was used at this pres- 

 sure. In the region up to 1.5 ^ no radiation from the hot cell could be 

 detected. Beyond 3 /t, for a pressure of 0.6 to i mm., the deflections 

 often amounted to several centimeters. On the other hand, nitrogen 

 shows strong lines at i ^, which would indicate that if nitrogen emits 

 bands on account of its thermal condition, then it must be at a higher 

 temperature than the cell. 



Cl6veite Gas. 

 In order to make the present work more complete, and from the fact 

 that cleveite gas (helium) is one of the few gases known to have emis- 

 sion bands in the infra-red, the curves (fig. 142) found by Runge and 

 Paschen,' are reproduced in this paper. They found the intensities of 





these lines to vary for diflferent samples of the gas. The band at 1,117 /* 

 gave a deflection of about 26 cm. The band at 2.040 /x is of interest, since 

 it coincides with the value computed from the spectral series formula. 

 Beyond 2 fx, as in the present work, no emission bands were found. The 

 continuous spectrum from 2.5 /x to 6 yu, is due to the radiation from the 

 hot cell, the depressions in the curve being due to atmospheric absorp- 

 tion bands of CO2 and HoO. 



Water Vapor (H2O). 



After finding the impurity band at 4.75 /x in all the gases except 

 hydrogen, and knowing that water has an absorption band at this point, 

 a special attachment was provided to introduce water into the pre- 

 viously exhausted pump and vacuum-tube. 



The emission spectrum was found for various pressures, up to 3 mm.. 



'Runge & Paschen : Astrophys. Jour., 3, p. 4, 1896. 



