58 Mr Arthur Schuster, On the passage of [Apr, 28, 



other part of the tube. Wullner has proved that the same holds 

 for oxygen. I have been able to confirm it in the case of carbonic 

 oxide and most likely in the case of chlorine. In the latter case 

 1 have not been able as yet to decide whether the observed 

 difference is not due to a mere difference in temperature. In all 

 the other cases it is certain that no differences in temperature or 

 pressure can produce the change in the spectrum. There is no 

 more doubt that the different spectra of one and the same gas 

 are produced by structural differences in the vibrating molecule. 

 These structural differences are produced in most, if not in all 

 cases, by a difference in the number of atoms making up the 

 molecule. If it can be shewn therefore that no mere difference 

 in temperature or pressure can produce the molecular structure 

 which gives rise to the spectrum at the negative pole, it must 

 be due to chemical or other causes operating at the negative 

 pole only of a vacuum tube. This proof I propose to give. 



In a great many cases the introduction of a Leyden Jar alters 

 the character of the spectrum. In all these gases, even without the 

 Jar, traces of the Jar-spectrum are seen in the negative pole. If 

 therefore temperature is the only operative cause, the temperature 

 at the negative pole must be intermediate between that in the 

 capillary part with and without the Jar, for without the Jar the 

 traces of the Jar-spectrum, if seen at all, are seen much weaker in 

 the capillary part than at the negative pole. By a suitable 

 intensity of discharge, a state can be established in which the Jar- 

 Spectrum is seen in the capillary part of a Geissler-tube, while the 

 wide part of the tube still retains the spectrum of the weaker 

 discharge. I have examined such a tube containing nitrogen, and 

 projected the image of the neck of the capillary part on the slit of 

 a spectroscope by means of a lens. I could see in the spectroscope 

 at the same time the spectrum of the capillary part and that 

 of the wide part. The appearance was as follows. The capillary 

 part shewed the line- spectrum of nitrogen, and the lines gradually 

 faded away towards the wide part. On the other hand, the wide 

 part shewed the bands of nitrogen which gradually faded away 

 towards the capillary part. The intermediate part was therefore 

 filled up with the longest lines of the line-spectrum aud the 

 longest bands of the band-spectrum. One cross-section could be 

 found, in which just those lines of the Jar- spectrum, which are seen 

 at the negative pole, were pi'esent. If temperature was the only 

 operative cause, the bands characteristic of the negative pole ought 

 to be observed at that point, but neither at that nor any other point 

 intermediate between the line and band-spectrum could the spec- 

 trum of the negative pole be seen. This can be observed through a 

 wide range of pressure, so that the pressure also may be eliminated 

 as operative cause. If it is thus conclusively established that 



