ON OUR KNOWLEDGE OF SPECTRUM ANALYSIS. 261 



Tliis experiment might not be considered to be altogether conclusive, 

 as a mere relative increase of intensity in the violet end by an increase 

 of temperature might not be considered sufficient evidence for such a 

 wide distinction. It was, therefore, thought better to discuss these 

 spectra together, and not to separate them, for there is no donbt that in 

 the vast majority of cases they appear as one whole and not as two dis- 

 tinct spectra. It might, however, be adduced in support of Pliicker and 

 Hittorf's opinion that the general aspect of the spectrum in the green is 

 certainly that of two overlapping spectra. Wiillner (' Wied. Ann.,' viii. p. 

 590), has described the gradual changes seen in a nitrogen tube having 

 a very fine capillary bore, when the pressure is gradnally reduced to a 

 very small amount. Owing to the increase of temperature the spectrum 

 gradually changes into a line spectrum, which is essentially the same as 

 the well-known line spectrum of nitrogen. Wiillner's results will be dis- 

 cussed in a separate report. 



The Spedrum of the Negative Glow. — The glow which surrounds the 

 negative electrode in an exhausted tube shows in many gases a spectrum, 

 which, as a rule, is not seen in any other part of the tube. In niti'ogen 

 this spectrum has often been observed since v. de Willigen first drew 

 attention to it, and was recently mapped by Angstroin and Thalen. It is 

 a channelled-space spectrum, fading away towards the blue. The bands 

 partially overlap some of the bands which are seen in the spectrum of the 

 positive discharge, so that with low dispersive powers it might seem as if 

 in the negative glow some of the ordinary bands were greatly increased 

 in intensity. But in reality a new series of bands is added at the nega- 

 tive pole, as will be seen with a good spectroscope, even if one prism only 

 be used. The ordinary spectrum of the positive discharge no doubt is 

 also present, though weak, in the negative glow, and often traces of the 

 spectrum of the negative pole are seen in the positive discharge ; but there 

 can be no doubt that we have to deal with two distinct spectra, although 

 it may not be easy to separate them entirelj^ When the pressure is much 

 reduced the negative glow gradually extends into the whole tube, and the 

 spectrum is then well seen in the capillary part. 



Discission on the Chemical Origin of the above Spectra. — Some discus- 

 sion has taken place on the chemical origin of the spectrum seen in the 

 positive discharge. In the year 1872 the writer of the present Report 

 ('Proc. R. Soc.,' XX. p. 482), described some experiments, in which he 

 showed that when sodium is heated in a nitrogen tube the band spectrum 

 disappears, and is replaced by a series of lines which he thought belonged 

 to nitrogen. He drew the conclusion from his experiment that the bands 

 were due to an impurity of an oxide of nitrogen. It has since, however, 

 been shown, especially by Salet, that the disappearance of the bands is 

 due to another cause, and that the line spectrum which appears on heating 

 the sodium is not due to nitrogen. Salet also showed how, with proper 

 precautions, sodium may be heated in a tube containing nitrogen without 

 destroying the band spectrum, and he has therefore furnished the proof 

 that this spectrum is really due to nitrogen, and not to an oxide. Wiillner 

 has also come to the same conclusion. Angstrom and Thalen, however, 

 in their joint paper, support the opinion that the spectrum is that of 

 some oxide of nitrogen. They try to support this view by an experiment 

 showing that when the brush discharge from a Holtz machine is observed 

 in atmospheric air, or when the ordinaiy discharge of a coil is sent through 

 rarefied air, the band spectrum is seen, and that at the same time the 



