322 Dr. A. Schuster. Experiments on the 



which an electric current is passing are always in a state of rapid 

 dissociation. Thus I do not know of a single case in which the gas 

 during the glow discharge does not show two, or generally even three, 

 distinct spectra in the same or in different parts of the tnbe. This 

 dissociation seems especially strong in the negative glow. The 

 characteristic spectrum which gases, as a rule, show in the neigh- 

 bourhood of the negative electrode bears evidence of belonging to a 

 complex molecule, but at the same time a high temperature line 

 spectrum is always seen to overlap the bands. I have shown some 

 time ago that the spectrum of the negative pole cannot be explained 

 by mere peculiarities of temperature, and, further, that it must be 

 due to the formation of a distinct molecule, as the spectrum keeps its 

 original place even a short time after the current has been reversed. 



As the subject is one that is not much studied, I may be permitted 

 to give here a short account of the spectral phenomena seen in tubes 

 filled with different gases. 



Nitrogen. This gas shows in the positive half a well-known com- 

 plicated band spectrum. The glow shows peculiar bands, but the 

 strongest nitrogen lines are also always visible, and it is generally 

 only there that the lines are seen. The capillary part will only show 

 lines when its bore is very small, as, for instance, when a piece of 

 thermometer tubing is used. As the bands of the positive half are 

 also seen in the glow, this region shows the superposition of three 

 distinct spectra. There is at present no longer any doubt that this 

 means the superposition of three distinct sets of molecules. 



Oxygen shows phenomena exactly corresponding to those in 

 nitrogen. The positive part as regards character shows what I liaye 

 called the complex line spectrum, which stands in a remarkable way 

 between a line and a band spectrum. The negative glow shows 

 characteristic bands, and at the same time both the spectrum of the 

 positive pole and traces of the high temperature line spectrum of 

 oxygen. Here, again, we have three distinct sorts of molecules in the 

 glow. 



The phenomena in hydrogen are not as yet so well ascertained. The 

 positive half and also the glow show generally a band spectrum 

 which now is generally ascribed to hydrogen ; but, at the same time, 

 the lines are very strong. I do not know whether a peculiar spectrum 

 has ever been noticed or looked for at the negative pole, but, owing to 

 the complexity of the band spectrum, it would be rather difficult to 

 discover. No gas shows greater difficulties of purification, and there- 

 fore the spectroscopic appearances are much more doubtful in this 

 than in the other gases. 



Carbon compounds show their characteristic spectra whenever they 

 are not very rapidly decomposed ; but at the same time they always 

 show at least in the negative glow the true carbon spectrum. I have 



