Intelligence and Miscellaneous Articles. 397 



acetylene mixed with a sufficient quantity of hydrogen remains un- 

 altered in a Geissler's tube, so that a decomposition, if produced, is 

 always accompanied by a corresponding combination, there is nothing 

 to prevent us from admitting that the hydrogen-spectrum No. 2 

 b elongs to acetylene. 



I pass now to the third of the spectra which M. Wiillner thinks 

 he has found for hydrogen. This spectrum, which would be quite 

 new if it belonged to the gas in question, is in all probability only 

 that of sulphur. This, I believe, is most positively demonstrated by 

 the following Table, which contains the wave-lengths of the vapour 

 of sulphur and those determined by M. Wiillner for this third spec- 

 trum of hydrogen. The differences met with in the two series are 

 easily explained by this consideration — that the wave-lengths cited, 

 both for sulphur and for M. Wiillner's spectrum, have not all the 

 accuracy desirable. Sulphur- 



Supposed Spectrum of Hydrogen. ,P. „ " m * 



1. Group of three lines; the middle line .... 5647 1 5645 



[5613 

 [5474 



2. Group of three lines ; the middle line .... 5469 i 5451 



[5432 



3. Group of two lines ; the second line 5334 j ^45 



4. Group of two series ; the first line 5221 I ^07 



[5027 



5. Group of three lines ; the middle line .... 5015 < 5013 



[4994 



6. Group of more than six lines; the middle line. 4930 4926 



I persist, therefore, in the opinion that hydrogen has only one 

 spectrum — the one found in the light of the sun and of the stars. 



Beside the known spectrum of oxygen, M. Wiillner has observed 

 two new ones, which, for brevity, we will designate by the numbers 

 2 and 3. According to the description given of No. 2, it is com- 

 posed principally of four shaded bands with clean edges on the side 

 towards the red field of the spectrum. In order to obtain a more 

 exact idea of the position of the bands, I constructed the spectrum 

 itself with the aid of the minima of deviation given by M. Wiillner. 

 I found that it presented much analogy with the spectrum of the 

 oxide of carbon. I afterwards determined the wave-lengths of the 

 four bands, by construction and with the aid of the wave-lengths 

 already calculated by M. Wiillner. The following Table gives the 

 values obtained, and the wave-lengths for the oxide of carbon : — 



Supposed Spectrum of Oxygen, No. 2. oL^f^bST 



1. Firsthand 5620 5609 



2. Second band 5203 5196 



3. Third band 4835 4834 



4. Fourth band 4506 4510 



