under the Influence of the Electric Discharge. 423 



tribution of the free electricity on the walls of the tube on the 

 free current-electricity. The wall, turned towards a positive 

 discharge, then naturally plays the part of a negative electrode. 

 In what way the discharges issuing from the two electrodes 

 equalize themselves is still an open question. 



I have already shown *, and also mentioned above, that the 

 phenomena produced by the kathode-rays cannot be caused 

 by projected molecules. If we were even to assume that the 

 thickness of a layer of such projected particles amounted to 

 a whole centimetre, that it therefore possessed a mass rela- 

 tively great, yet, in order to produce the observed heating 

 of the tube above the negative electrode, it must possess velo- 

 cities of 100,000 metres or more. 



Gintlf and PulujJ seek to refer the phenomena to a pro- 

 jection of the substance of the electrodes ; but since with alu- 

 minium electrodes, for example, no decrease of the electrodes 

 is observed, the velocity must be immense. This view is 

 therefore scarcely tenable. Moreover all the other pheno- 

 mena observed may be explained by the propagation of 

 sether-waves of very great energy ; thus, for example, the 

 mutual influence of two rays may be explained by the pres- 

 sures occurring on the anterior surface of advancing waves §, 

 which in our case may be of considerable magnitude. 



Let us turn, lastly, to the results in spectrum-analysis which 

 may be drawn from the above investigation. In a previous 

 investigation || I have endeavoured to show, in continuation of 

 the investigations of Lockyer, Lecoq de Boisbaudran, and 

 others, that the band spectrum corresponds to the individual 

 molecule, the line spectrum to the separate atoms. The same 

 quantity of heat which is necessary to change the band spec- 

 trum into the line spectrum gives, therefore, a maximum value 

 for the heat of decomposition of the hydrogen molecules, of 

 which, besides the work expended in the dissociation, another 

 portion is employed in heating the gas. 



We find that about 128,000 gramme-calories must be commu- 

 nicated to 1 gramme of hydrogen of ordinary temperature in 

 order to separate it into its atoms. This number is to be re- 

 garded as only a rough approximation, since its determination 

 is affected by so many sources of error ; it must remain for 

 further investigations to determine this number more exactly. 



It corresponds to values such as we frequently have in 



* E. Wiedemann, Wied. Ami. ix. p. 157, 1880. 



t Separate Publication. 



% Puluj, Wiener Anzeiyer, p. 70, 1880. 



§ Maxwell, ' Electricity and Magnetism,' ii. p. 391. 



|| E. Wiedemann, Wied. Ann. v. p. 512, 1878. 



2H2 



