1881.] Note on the Spectrum of Carbonic Acid. 381 



difficulty, for the green hydrogen line was seen with great constancy 

 in high vacua, even if not the least trace of it could be remarked 

 when the carbonic acid spectrum could be seen brightly. I soon 

 recognised that the more the gas had been deprived of moisture, the 

 further had the exhaustion to be carried on to render the said line 

 visible again, but for a long time I could not get quite rid of it. By 

 repeatedly filling and exhausting the vacuum tube, in doing which the 

 gas naturally was only permitted to pass very slowly through the 

 phosphoric anhydride tube, and the spectrum tube was heated to the 

 highest degree possible, I at last arrived at such a degree of perfect 

 dryness as no more to give any trace of the fatal green line, not even 

 if the tube was exposed to high temperature ; this, indeed, always had 

 the effect of driving into the vacuum some matter adherent to the 

 sides of the glass, and causing the illumination of the tube, when this 

 had been so far exhausted as no longer to permit any current to pass, 

 to reappear. But the brilliant mercury spectrum seen under such cir- 

 cumstances showed nothing of a hydrogen line. Nevertheless, after 

 this exceedingly high degree of dryness had been obtained, when there 

 was some gas admitted into the spectrum tube and this again ex- 

 hausted, the fatal line was feebly seen, showing clearly that the gas 

 had not been entirely deprived of its moisture by its passing through 

 the system of drying apparatuses. I therefore finally only experi- 

 mented with such carbonic acid as had been in perpetual contact with 

 the phosphoric anhydride during at least twelve hours, and it was not 

 till now that I got rid of every trace of the hydrogen lines. When 

 making the experiments that I believe to be decisive, I always took 

 care to convince myself, by rarefying, not only that the gas had 

 remained perfectly dry, but also that the current had not introduced 

 any trace of hydrogen or hydrocarbon when passing through the tube ; 

 for both would have been discovered in the high vacuum by the 

 appearance of the green line, the spectrum of hydrogen being by no 

 means oppressed by that of carbon in such a vacuum, as I most clearly 

 observed when experimenting with olefine gas, which in a very rarefied 

 state shows from the beginning of the current's passage a very dis- 

 tinct spectrum of the first order of hydrogen, besides the one of carbon. 

 If the bands and lines observed by Swan really belonged to hydro- 

 carbon only they could not have appeared in our case, but they did 

 appear very distinctly and clearly, as soon as the density of the gas 

 became great enough to oblige the current to pass as a spark 

 through the vacuum tube ; and even not less brightly was Swan's 

 spectrum seen at much lower pressures by using a Leyden jar. The 

 capillary tube in which the said spectrum alone was seen was then 

 wholly filled by the luminous discharge ; no real spark appeared in 

 this case, while the wide parts of the spectrum tube showed 

 Watts' second carbon spectrum. It is these last-mentioned experi- 



2 d 2 



