between Anode and Kathode Spectra. 345 



Other gases which I have studied are the oxides of carbon. 

 So far as I can see they each give, with a continuous discharge, 

 but one spectrum in all parts of the tube, in the negative glow, 

 the capillary, the stria? and the anode, it is the same spectrum for 

 both carbonic oxide and carbon dioxide, and was faithfully 



described and figured by Angstrom and Thalen, in their Re- 

 cherches sur les spectres des metalloids before referred to, as the 

 spectrum of carbonic oxide. It consists of a number of channelled 

 bands degraded on the more refrangible side, and of a number of 

 weaker bands, almost equally diffuse on their two sides, in the 

 spaces between the degraded bands. These gases are very sensi- 

 tive to the kathode rays, and give a bright negative glow, by 

 which the course of the kathode rays may be readily traced. The 

 light in all parts of the tube is nearly white, and in the strise, 

 which are easily produced, there is no perceptible difference of 

 colour between the anode and kathode sides of the strise, as there 

 is in the strise of hydrogen and of nitrogen, nor such a well- 

 marked position of the maximum of luminosity. There was no 

 trace of the bands of the candle-flame spectrum in either of the 

 two gases in any part of the tube, with such a discharge as I have 

 used. 



Another carbon compound examined was cyanogen. The 

 observations were impeded by the rapid decomposition of the 

 compound, and the blackening of the tube, but I was able to 

 observe some effects of the discharge. When all carbonic acid 

 gas had been completely removed, the tube refilled with cyanogen 

 and the pressure reduced, on first passing the discharge strise 

 appeared fairly sharply defined and moving from the kathode 

 rather quickly, but as decomposition proceeded the striae became 

 very diffuse and ceased to move. After pumping out and refilling 

 and reducing the pressure to about 2 mm., the kathode glow was 

 fairly bright and the dark space defined. The spectrum of the 

 glow shewed a set of bright bands in the orange and yellow, which 

 appeared, by their number, character and arrangement, to be the 

 same as the bands in that region of the spectrum which are 

 developed in the flame of cyanogen burning in oxygen, though 

 I could not prove identity, not having such a flame at hand for 

 comparison. Then there were besides two sets of bright bands in 

 the blue and violet, which from their position and appearance 

 I had no doubt were the cyanogen groups beginning at about 

 X 4600 and 4215 respectively. At the anode no blue or violet 

 bands could be seen, only some continuous light in the orange and 

 on to green, with traces of the bands in that region that were 

 bright and well-defined in the kathode glow. At higher pressure 

 the anode light gave these same bands distinctly, though not 

 so bright as the kathode glow, while the blue and violet sets were 

 vol. xu. pt. v. 23 



