Dr. W. M. Watts on the Spectra of Carbon. 253 



A Geissler's tube enclosing marsh-gas gives y, S, £, and 6, 

 but the group e is not observed. This spectrum contains also 

 a line at 74, which may belong to the second carbon-spectrum. 



Carbonic oxide has been stated to yield the ordinary carbon- 

 spectrum when the induced spark is taken in the gas at the or- 

 dinary pressure. The discharge through a Geissler's tube, 

 however, exhibits an entirely new spectrum which contains none 

 of the ordinary carbon-lines. That this new spectrum is also 

 due to carbon itself is shown by the fact that it is obtained either 

 from a vacuum-tube enclosing carbonic oxide, or from one enclo- 

 sing oleliant gas*; and it becomes a question of much interest 

 to determine upon what conditions the production of one or the 

 other of these forms of the carbon-spectrum depends. Olefiant 

 gas is capable of yielding either spectrum. When the discharge 

 is passed through a tube containing olefiant gas of only a few 

 millimetres pressure, the spectrum drawn (fig. 2e) is obtained, 

 but the gas at the ordinary pressure yields the first form. In 

 order to determine at what pressure the second spectrum dis- 

 placed the first, a tube provided with platinum wires was con- 

 nected with the air-pump so that it could be exhausted, and by 

 means of a tap with a source of olefiant gas. It was also pro- 

 vided with a gauge-tube, by means of which the pressure could 

 be measured. When the pressure is about 12 millims., the 

 spark is violet and gives the carbon-spectrum of fig. 2; when 

 the pressure of the gas was gradually increased the spark became 

 blue; and at a pressure of about 100 millims. the spectrum 

 changed to that of the first form. When still more gas was ad- 

 mitted the spark became white, and carbon was rapidly sepa- 

 rated. 



Pliicker t has observed these lines of the second carbon-spec- 

 trum. In his earlier paper he describes them as lines belonging 

 to the compound carbonic acid; but in the paper published in 1865 

 he represents them as belonging to carbon itself. Fig. 2, a & b, 

 shows the observations of Pliicker, reduced from the drawing to 

 his paper in the Philosophical Transactions to the scale employed 

 throughout this paper. Fig. 2 a shows the spectrum obtained 

 from spectral tubes enclosing carbonic oxide of 32 millims. pres- 

 sure. A comparison of this spectrum with that of carbonic oxide 

 (fig. 2 c) and with that of olefiant gas (fig. 2e), shows that Pliicker 

 did not succeed in completely separating the two spectra. I 

 have, however, repeatedly obtained the second spectrum alone, 

 consisting of the bands h, j, k, I, m, n } and o, and exhibiting no 



* This curious difference in the spectra obtained from different carbon- 

 compounds was first noted by Dr. Roscoe, in a lecture delivered before the 

 Royal Institution in May 1864. 



\ Pogg. Ann. vol. cvii. (1859). Phil. Trans. 1865. 



