456 Dr. W. M. Watts on Carbon- Spectra. 



where dp is the variation of pressure, and d<f> the variation of the 

 thermodynamic function. 



In this equation dv, dt correspond to ^ Y , i/r 2 , dp, d<f> to W v ^ 2 , 

 and de corresponds to the potential energy of the purely mecha- 

 nical system. Our first proposition shows that, if dcf> = 0, -j- 



dv * 



has the same value as that of ■ L rr when there is no variation of 



dxp 



pressure, the interpretation of which is that the heat (measured 

 as work) necessary to increase the volume by unity at constant 

 pressure, is numerically equal to the product of the absolute 

 temperature into the increase of pressure required to raise the 

 temperature one degree when no heat is allowed to escape. (See 

 Maxwell's ' Heat/ p. 167.) In like manner the other proposi- 

 tions may be interpreted. 



LX. Note on Carbon- Spectra. By W. Marshall Watts, D.Sc, 



Physical- Science Master in the Giggleswick School*. 



IN the Philosophical Magazine for October 1869 I described 

 four different spectra as spectra of carbon. One of them 

 was the ordinary spectrum from hydrocarbon-names first de- 

 scribed by Swan; the second was the spectrum obtained from 

 vacuum-tubes enclosing carbonic oxide, carbonic anhydride, or 

 olefiant gas; the third was the spectrum of the Bessemer-flame, 

 and the fourth the spectrum of the high-tension spark in car- 

 bonic anhydride or carbonic oxide 



I have since shown (Phil. Mag. February 1873) that the 

 Bessemer-spectrum, instead of being a spectrum of carbon, is 

 the spectrum of manganese oxide ; and I have now to add 

 the result of recent observations which show that the second 

 spectrum also is due not to carbon itself, but to some oxide of 

 carbon. This spectrum was held to be a spectrum of carbon 

 because it was common to compounds of carbon with hydrogen 

 and with oxygen. Following the line of research indicated by 

 Schuster (Proc. Roy. Soc. June 1872), I have now found that 

 it is not given by spectral tubes enclosing olefiant gas if spe- 

 cial care be taken to exclude all trace of oxygen. A descrip- 

 tion of one experiment will suffice. The olefiant gas prepared 

 from alcohol and sulphuric acid passed first through a wash- 

 ing-bottle containing sulphuric acid, then through a second 

 bottle filled with solid caustic potash. This bottle was also 

 provided with a separating-funnel, the tube of which passed 

 through a third hole in the cork nearly to the bottom. When 

 the olefiant gas had passed through the whole apparatus for 

 * Communicated by the Author. 



