Dr. W. M. Watts on the Spectra of Carbon. 255 



line first becomes visible ; then the lithium-line is added, and 

 gradually the lines of the Bessemer-spectrum, increasing in bril- 

 liancy to the end of the blow. 



The spiegel-spectrum, as I have pointed out, is only the Bes- 

 semer-spectrum in which some of the lines are still further in- 

 creased in brilliancy, and is doubtless due to the highest tempe- 

 rature of all ; for we have the hot carbon of the molten spiegeleisen 

 burnt by the intensely heated oxygen absorbed by the liquid 

 steel. The spiegel-spectrum is occasionally identical with the 

 ordinary Bessemer-spectrum, when, namely (as shown by the 

 spectroscope and by the analysis of the steel), the blast has been 

 stopped somewhat short of the proper point. The effect of an 

 increase of temperature is thus to split up the Bessemer-spectrum 

 into groups of lines, in each of which the brightest line is the 

 most refrangible — an aspect which is exactly the reverse of that 

 so noticeable in the ordinary carbon-spectrum, where each group 

 has its strongest line on the left hand. 



& fourth spectrum, also probably due to incandescent carbon, 

 is obtained from the induced spark in either carbonic oxide or 

 carbonic anhydride when a Leyden jar is included in the circuit, 

 and is represented in fig. 4. It is one of the spectra termed 

 by Pliicker " spectra of the second order, " consisting, not of 

 bands, but of sharply defined lines, frequently in pairs. It 

 has been already stated that the induction-spark (without con- 

 denser) gives in carbonic oxide the carbon-spectrum No. 1, and 

 in carbonic anhydride a continuous spectrum. With a suffi- 

 ciently large condenser the spectrum obtained from carbonic 

 oxide is identical with that obtained from carbonic anhydride, as 

 will be seen on comparing fig. 4 a (spectrum of carbonic oxide) 

 with fig. 4Z> (spectrum of carbonic anhydride). The carbonic 

 oxide was prepared from potassium ferrocyanide and well washed 

 with caustic potash. The spectrum obtained from air under 

 similar conditions is subjoined for the sake of comparison. The 



carbon double band < J^t is at first sight identical with the 



double band in the air-spectrum. If, however, while the spark 

 continues to pass, the carbonic anhydride be blown out of the 

 discharge-tube and replaced by air, it is distinctly seen that the 

 two are not coincident. The left-hand nitrogen-line is slightly 

 more refrangible than the left-hand carbon-line ; the right-hand 

 members are (with one prism) apparently coincident. 



The continuous spectrum obtained by the discharge of an in- 

 duction-coil in carbonic anhydride may be converted into this 

 fourth carbon-spectrum, either by increasing the electric conden- 

 sation as described above, or by increasing the density of the gas. 

 Carbonic anhydride in the compression-apparatus which I have 



