Second Spectrum of Hydrogen. 337 



that Berthelot's third spectrum belongs to acetylene; he has 

 only inferred, from the processes of discharge in the appear- 

 ance of this spectrum on the one hand, and of the ordinary 

 spectrum of hydrogen on the other hand, that we have to do 

 in the first case with a compound of hydrogen, " soit avec 

 lui-meme soit avec un corps etranger;" and then, on account 

 of the stability of the mixture of acetylene and hydrogen as 

 shown by Berthelot, he takes the latter alternative as the 

 more probable. If, now, in consequence of evidence already 

 brought forward, and further on account of evidence still to be 

 adduced, the assumption of a compound with a foreign sub- 

 stance appears to be little probable, there is no reason that 

 we should lose sight of the possibility of the other alternative 

 suggested by Angstrom, according to which the spectrum is 

 to be ascribed to a compound of hydrogen with itself. On the 

 contrary there are, as we shall presently see, reasons which 

 are decidedly favourable to such an assumption. 



If further, in Q the second place, we remember that the 

 investigations of Angstrom and Thalen, which treat, amongst 

 other things, of the spectroscopic relations of carbon com- 

 pounds, and give the first rational explanation of the spectro- 

 scopic phenomena of hydrocarbons, were not published until 

 four years later, viz. in the year 1874, and moreover, in conse- 



o 



quence of Angstrom's death, were not so complete as was 



o 



intended, we have the explanation of the fact that Angstrom 

 did not subject his first theory of the second hydrogen spec- 

 trum to further proof. 



Berthelot's theory receives a confirmation, which, at first 

 sight, seems of great weight, from the researches of Salet *• 

 If, in fact, Pliicker's spectrum really belongs to acetylene, 

 then this spectrum cannot appear in a hydrogen-tube which is 

 free from all impurity of carbon compounds. In order to test 

 this, Salet allowed a current of pure dry oxygen to circulate 

 for a long time through the tube to be used before it was filled 

 w T ith hydrogen, the tube being all the time heated to a low 

 red-heat. In this way it was believed all traces of carbon 

 compounds must have been removed. Further, in order to 

 eliminate all possible influence of the electrodes, they were 

 omitted altogether, and instead the wide parts of the tube w r ere 

 coated with tinfoil : when these were connected with the poles 

 of the induction-coil, or with the combs of the Holt z machine, 

 the gas became luminous. Under these circumstances spec- 

 troscopic examination of the tube showed only the three 

 principal lines of hydrogen without any trace of the second 

 * Salet, Spectres des Metalloides (Paris, 1872), p. 17. 



