342 M. E. Goldstein on Spectra of Gases* 



the jar is charged by an induction-coil the current-intensities 

 of which are variable according to the quicker or slower action 

 of its current-break. If the band-spectrum was just obtained 

 with a quick vibration of the hammer, it sufficed to force this 

 to slower action by increasing its weight, in order immedi- 

 ately to change the spectrum into that of the second order. 

 If the (small) density of a narrow tube which yields a band- 

 spectrum even with jar- discharges is left constant, and the 

 exterior spark-distance of the inserted jar is increased, the 

 band-spectrum can be converted into one which is composed 

 of spectra of both orders, and finally into the pure line-spec- 

 trum. The influence of the change in the current-intensity 

 can be controlled in a striking manner, if the density is kept 

 constant, by the enlargements of the hydrogen-lines ; in a tube 

 consisting of parts of different widths the hydrogen-lines in 

 jar-discharges are all the more enlarged the narrower that part 

 of the tube is which surrounds the gas. In one and the same 

 part of the tube the lines are the broader the greater the distance 

 in air is which is inserted in the arc. If the same current 

 passes through several tubes of different, (but also constant) 

 density, the lines are broader in a very narrow tube with 

 strongly rarefied gas than in a wider one with gas of relatively 

 high density. 



If into the current which enlarges the line of hydrogen a 

 moist thread be inserted, the breadth of the lines gets smaller; 

 and by sufficient length of the thread they are restored to per- 

 fect sharpness. 



With certain low pressures and in certain experiments the 

 lines of pure hydrogen, in spite of large spark-distances inserted, 

 were yet only very little enlarged by jar-discharges. It was to 

 be proved whether in reality there are limits of density beyond 

 which a change of the spectrum is no longer caused by tempe- 

 rature. 



As with the use of moist air for discharges of the simple in- 

 duction -current the lines of hydrogen appear simultaneously 

 with the bands of nitrogen, this indicates a heat of the conti- 

 nuous discharge of the latter which equals the spark-temperature 

 of hydrogen. It was therefore to be expected that, from the 

 strong discharge of sparks with line-spectrum in nitrogen, a 

 heat would result far surpassing the spark-temperature of the 

 former gas ; and the ignition of hydrogen in the nitrogen spark 

 necessarily exposed it to far higher temperatures and their 

 effects than the spark- discharge in pure hydrogen itself. 



In reality the lines of hydrogen which was mixed with nitro- 

 gen were far more dim than those of the pure gas at the same 

 pressure. 



