Spectra in Compound Gases. 729 



hydrogen was very brilliant, and was accompanied by some- 

 what faint negative bands of nitrogen. At first the Swan 

 spectrum was seen for a few seconds at the anode, but it was 

 rapidly replaced by carbonic oxide. There were also very 

 faint traces of a number of bands in the violet, most pro- 

 bably constituting part of the positive pole spectrum of 

 nitrogen. There appeared to be no tendency to deposit 

 carbon on the walls of the discharge-tube. 



Thus it seems as though the electropositive amido-group 

 had a tendency to go to the cathode, and the benzene nucleus 

 to the anode. This would explain the presence of the spectra 

 of hydrogen and. nitrogen at the negative pole, and of carbonic 

 oxide at the positive pole. 



(vii.) Ethyl Ether. — When the coil was first turned on the 

 Swan spectrum was bright and the carbonic oxide spectrum 

 faint at the positive pole, whilst at the cathode the converse 

 was true. After a few seconds the Swan spectrum at the 

 anode was replaced by carbonic oxide bands, whilst at the 

 cathode the ordinary spectrum of hydrogen became visible. 

 No carbon was deposited on the walls of the tube, but, in 

 spite of this, decomposition apparently took place, for the 

 passage of the discharge caused a considerable permanent 

 rise of pressure. 



(viii.) Methyl Sulphide. — Initially both poles showed Swan 

 bands, but these were rapidly replaced at the cathode by 

 carbonic oxide and the four-line spectrum of hydrogen. 

 Later on sulphur bands appeared at the anode, and after 

 remaining fairly bright for some minutes at length gave place 

 to the carbonic oxide spectrum. Decomposition took place 

 fairly readily and carbon was deposited on the sides of the tube. 

 Thus in methyl sulphide sulphur shows a preference for the 

 anode just as in the case of hydrogen sulphide and carbon 

 disulphide. 



(ix.) Chloroform Vapour in a very short tube. — A tube was 

 used in which the electrodes were only about 1*5 cm. apart. 

 With pressures of the order of 0*5 cm. and a weak discharge 

 perfectly normal results were obtained. When, however, 

 the pressure was taken down to about half a millimetre, and 

 a heavy discharge was sent through the tube, the discharge 

 became completely oscillatory, in spite of the "point and 

 plane " spark-gap in use. The anode and cathode could be 

 distinguished no longer, whilst the positive column practically 

 disappeared. In these circumstances chlorine lines appeared 

 throughout the tube, and very little could be seen of the 

 carbonic oxide bands which are normally present in the 

 positive column. 



