99 
reason is that the bands are found in the spectrum of the sun. It is dif- 
ficult to believe that a carbon nitrogen compound exists at so high a tem- 
perature, or even at the temperature of the are. Another reason is that 
the strength of the bands does not vary much when the quantity of N in 
the arc is varied. 
All the investigations on this subject show that a mere trace of N is 
sufficient to bring out the cyanogen bands quite clearly while the carbon 
bands are weak. Flooding the arc with N does not greatly increase the 
strength of the cyanogen bands or weaken the carbon bands. The follow- 
ing experiment emphasizes this point: 
Mercuric cyanide (30 gms.) was placed in a three-bulb combustion tube. 
One end of the tube was sealed and the other connected by a rubber 
tube to the hollow carbon.of the inclosed are. The tube was heated by 
gas burners, and as soon as the cyanogen had displaced all the air in the 
globe the are was started. The three cyanogen bands were very slightly, 
if any, stronger than in the are in air, and not very much stronger than 
when treated carbons were used with CO, streaming into the are ana 
globe. Yet the quantity of cyanogen in the arc must have been hundreds 
of times greater in the first case. That the two spectra shouid be at all 
comparable in intensity is not in agreement with our knowledge of spectra 
in general. 
The author inclines to the view that the so-called cyanogen bands are 
due to carbon in the presence of nitrogen, but not combined with it. It 
is well known that the spectrum of an element is sometimes greatly 
changed in the presence of another element, though there be no tendency 
for the elements to combine. The statement need not be confined to spec- 
tra. A very familiar instance is the addition of MnO, to KCl1O, in the 
generation of O. The temperature at which the O is given off is greatly 
reduced, although the MnO, does not decompose nor does it combine with 
the K. 
It has been shown that cyanogen is produced by an are in air or N. 
However, the compound may be formed in the outer cooler portions of 
the arc and_not in the central hot region, where the cyanogen bands are 
strongest. The fading out of these bands in the outer regions is more rapid 
than would be expected if cyanogen were a stable compound in the are. 
And to account for their production by a mere trace of N there must be a 
decided tendency for the compound to form. But experiments show that 
e€yanogen may be decomposed in the are. 
