betufeen Anode and Kathode SpecircL 349 



spectrum is chiefly ultra-violet, so that it is a priori not im- 

 probable that it would give no visible fluorescence. It is also not 

 improbable that the ions are furnished by dissociation of the 

 elements, and the facts that water vapour gives no anode light, 

 hydrochloric acid a continuous spectrum at the anode, lend con- 

 firmation to the supposition that oxygen furnishes the positive 

 ions in the former, chlorine in the latter medium. 



The identity of the spectra given by the two oxides of carbon 

 suggests that the ions of carbonic acid gas are furnished by dis- 

 sociation of carbonic oxide and oxygen. It is true that I have not 

 been able to trace the presence of oxygen at the kathode as I was 

 able to do in water vapour, but considering that in the latter case 

 it was only blue lines of oxygen that were seen, and that they 

 could not have been detected but for the fact that hydrogen gives 

 no light in that region, it is not surprising if the very brilliant 

 spectrum of carbonic oxide overpowered the oxygen. 



It has been maintained that the common spectrum of the two 

 oxides of carbon is emitted by carbon, but the fact that cyanogen, 

 when all carbonic acid from the glass has been well pumped out, 

 does not shew the carbonic oxide spectrum, is a strong reason for 

 concluding that molecules of carbonic oxide are really the origin 

 of the vibrations producing the spectrum in question. 



In conclusion, the examination of the kathode glow, very 

 incomplete as it is, has suggested to me that the explanation of 

 the solar chromosphere and corona will be found in regarding 

 them as a huge kathode glow. 



G. D. L. 



