1877.] electricity through gases. 59 



special causes must be at operation at the negative pole of a dis- 

 charge through a gas, it is no more astonishing that traces of the 

 simpler spectrum of the jar-discharge are seen at the negative pole. 

 For it is probable that the combination which gives rise to the 

 spectrum of the negative pole is due to the action of the simple 

 atom giving the line-spectrum, which is set free at the electrode. 

 I wish to draw attention to the fact that ozone is chiefly formed 

 at the negative pole, and that though we know that ozone is 

 destroyed by temperature, we have no right to suppose that tem- 

 perature will destroy all similar combinations if formed. 



There is one observation which I believe to be of sufficient 

 interest to be recorded, because it shews that there is no real 

 difference between the spark-discharge at ordinary pressures and 

 what we call a continuous discharge in gases. If a spark is sent 

 through oxygen at ordinary pressures, the line-spectrum is seen, but 

 if the intensity of the spark is reduced below a certain point, the 

 spark takes a yellow colour, and a purple point is seen at the 

 negative pole. The yellow spark shews a continuous spectrum, while 

 the purple point shews the spectrum of the negative pole. The 

 whole spark has in the spectroscope the same appearance as the 

 continuous discharge in a Geissler tube. 



It is no argument against the view which I have brought for- 

 ward, that the spectrum of the negative pole is seen under certain 

 circumstances at places removed from the metallic electrode. 

 Faint traces of the spectrum are sometimes seen throughout the 

 tube, and under very great exhaustion it is always seen even in the 

 capillary part. The diffusion of gases sufficiently accounts for 

 this fact, which would be difficult to explain if we did not assume 

 that the spectrum is due to a distinct molecular combination. 



Again, Mr Goldstein has observed that at all places where 

 negative electricity comes out of a narrow into a wide part of a 

 tube, a glow is seen which in all its properties resembles the glow 

 surrounding the negative electrode, and the spectrum of the 

 negative pole is seen at that place. The causes which produce the 

 combination give rise to that spectrum, and must therefore be in 

 operation at all points in which there is rapid increase in cross- 

 section in the direction of the negative current. 



The most difficult part of the problem of the conductivity of 

 electricity through gases is the passage of electricity from the 

 electrode to the gas. The important part which the electrode 

 plays has long been known. The negative electrode especially is 

 disintegrated, and parts of it are deposited on the glass-wall 

 surrounding the electrode. Vacuum tubes, owing to long-continued 

 use, shew sometimes remarkable changes. Mr Goldstein has found 

 that these changes are due to the absorption of the gas by the 



5—2 



