422 Trowbridge — Spectra of Gases. 



product of disintegration of radium ; a transmutation, so to 

 speak ; and is not a result of the electrical stimulus in the 

 environment of glass or quartz a stimulus which may bring to 

 light the helium which has refused to manifest itself by chem- 

 ical analysis 1 



In general it may be said that the greater the conductibility 

 of the volatilization, products either from the walls of the tubes 

 or from the metallic terminals determine the occurrence of the 

 spectral lines or bands. The spectrum, for instance, of silica 

 completely masks the spectrum of the iron terminals w T hen the 

 latter are placed not more than five millimeters apart. When 

 the terminals are of different metals the spectrum of the more 

 volatilizable metal predominates : or more strictly, the spectrum 

 of the better conducting vapor. 



Another striking fact brought to light by such discharges in 

 capillaries is the reversal of many of the spectral lines on 

 broad bands. The broadening of the lines of the metals is 

 generally toward the red end of the spectrnm. The quantity 

 of the discharge appears to be the important factor in deter- 

 mining the character of the spectra ; electromotive force, jper se, 

 does not give new lines which can be detected by photography. 

 The effect of high electromotive force begins to be evident at 

 high exhaustions and then only in producing cathode and 

 X-rays. 



This latter fact can be well shown by a Tesla coil actuated 

 by a Cooper-Hewitt mercury interrupter such as was employed 

 by Dr. G. W. Pierce (Proc. Am. Acad., 1904). With a suit- 

 able step-up transformer, in connection with such an inter- 

 rupter, I have studied the spectrum of hydrogen, and have not 

 obtained a spectrum which differed from the one obtained by 

 the same amount of energy with a lower voltage. The high 

 voltage ranged from 100,000 volts to 3,000,000. 



The broadening of metallic lines seems to indicate an oxidiza- 

 tion. One can conceive of a loading of the metallic molecule 

 by various degrees of oxidization which leads to a broadening 

 towards the red end of the spectrum, or in other words to 

 longer wave lengths, and an unloading due to dissociation 

 which leaves the molecule free to emit shorter wave lengths. 

 That an oxidization results from a discharge of electricity in 

 glass or quartz tubes filled even with apparently dry hydrogen 

 seems to me to be evident from my experiments. The unavoid- 

 able presence of water-vapor in glass, and I may add, in quartz 

 tubes, lends color to this oxidization theory ; this vapor is dis- 

 sociated by the electric current, the oxygen, set free, combines 

 with the molecules of the metals, or with the molecules of 

 silica and its metallic impurities. 



The following experiment illustrates this oxidization : 



