Harritey— Occurrence of Cyanogen Compounds n Coal-Gas. 291 
“marking in the continuous spectrum, or band,” 4215, 4208, and 
4196: opposite to these are the wave-lengths of three bands 
measured by Kayser and Runge, 4215-26, 4208-24, 4196-05. 
There is no statement to the effect that these are cyanogen 
bands or that they were believed to be such; moreover, in the 
general description of the spectrum on p. 170, it is expressly 
stated, “‘ All the principal bands observed are probably due partly 
to carbon and partly to what generally is considered to be the 
cyanogen spectrum.” 
It will be noticed that the word probably is used, and the bands 
_ named are designated as those generally considered as the cyanogen 
spectrum. Moreover, it is to the principal bands that this remark 
is applied. ‘This clearly indicates what was actually the case, 
namely, that I had not concluded they were carbon bands, and 
had not believed them to be cyanogen bands. In point of fact, 
as two subsequent publications prove,' there were facts known to 
me at that time which rendered it difficult to believe in the exist- 
ence of a distinct and individual cyanogen spectrum. Photographs 
of a number of flame spectra, including that of cyanogen-gas 
and of ordinary coal-gas had been obtained some years previously, 
namely, in 1882; but in view of the facts that some of the 
eyanogen bands could be obtained by passing sparks between 
graphite electrodes in air, when moistened with strong solutions 
of such chlorides as those of zine and calcium, and could not be 
obtained when the graphite was moistened with cyanides, such as 
those of potassium and of mercury, it appeared to be not improb- 
able that these were in reality a modification of the carbon 
spectrum. At the same time it was known that imperfectly 
purified coal-gas might contain cyanogen compounds, and the 
combustion of these would give the banded spectrum or a portion 
of it, peculiar to the cyanogen flame. Even the combustion of 
carbon at a high temperature in the presence of nitrogen in air 
might be expected to do so. 
It will thus, I think, appear evident that my views at the time 
were clearly stated, but they have been imperfectly understood, 
and to some extent misinterpreted. Had I been aware of the 
1 Qn Variations observed in the Spectra of Carbon Electrodes,’’ Proc. Roy. Soc., 
vol. 55, p. 344, 1894, and “ On the Spectrum of Cyanogen as produced and modified by 
Spark Discharges,’’ Proc. Roy. Soc., vol. 40, p. 216, 1896. 
