338 M. E. Goldstein on Spectra of Gases. 



which the mirror cannot widen. The spectrum, however, re- 

 mains the same. If under higher pressure it was a band-spec- 

 trum, it will remain so. 



If the intensity of the current decreases, the continuous dis- 

 charge often alternates with sparks— as, for instance, if the clo- 

 sing of the inducing current happens to be bad or too short. 

 Artificial strengthening of the current does away with these 

 sparks; the introduction of moist resistances increases their 

 number. The spectra do not change if this takes place. Very 

 often these sparks pass at the beginning of the discharge, or 

 after having changed the direction of the current. The band- 

 spectrum of air, of carbonic oxide, &c. in all these cases is 

 permanent. 



(As the introduction of a spark in the external air generates 

 in the rotating mirror another image of the discharge, which 

 to the eye is superposed on the continuous discharge, it is not 

 advisable to do away with the closing- current by means of a 

 spark in the air. This is best done by means of tubes which 

 only allow the opening-current to pass.) 



If the vacuum-tube contains moist air, the spectrum shows 

 the lines of hydrogen together with the bands of air. Accord- 

 ing to "Wullner's theory, only the continuous discharge can 

 produce the band-spectrum of air, while only the disruptive 

 discharge can produce the line-spectrum of hydrogen. The 

 rotating mirror should therefore show sharply defined images of 

 the tube due to the disruptive discharge of hydrogen, and at the 

 same time broad fields due to the continuous discharge through 

 air. This, however, is not the case ; the mirror only shows uni- 

 form fields, proving that there is only a continuous discharge 

 passing, and that therefore the hydrogen lines can be produced 

 by a continuous discharge. 



A tube filled with hydrogen under low pressure mixed with a 

 convenient quautity of oxygen gives out a magnificent crimson 

 light when the induction-current passes. No bands are seen; 

 neither, as a rule, are any lines visible which are not due to 

 hydrogen. (In some cases two greenish yellow lines could not 

 be eliminated.) The rotating mirror shows that the discharge 

 is continuous. 



If we slacken the speed of the interruptor of a Ruhm- 

 korfFs coil, the intensity of the light given out by a vacuum- 

 tube increases up to a certain point. Looked at through the 

 spectroscope, the hydrogen-lines are seen to flash in with great 

 intensity with every opening-current. The rotating mirror 

 shows that at the same time the long continuous band increases 

 in intensity. The discharge, therefore, is still continuous. It 

 can be directly shown that band- and line-spectra may be 



