and -other Spectra with the Spectrum of the Aurora. 263 



front of a more remote range. On this point, too, note the 

 peculiarities of the red line, which (and, as I find, the green line 

 also) is coincident with, or very close to, telluric bands or groups 

 of lines in the solar spectrum (see Plate III. fig. .2). 



I think we may also claim the continuous spectrum in the 

 aurora in further proof of water-vapour (see Plate V. spectrum 7). 

 The continuous spectrum of the aurora is also to my eye more 

 local and dense in the spectroscope than the usual spectrum seen 

 between the lines or bands in gases. 



B. The question of the violet-pole spectrum. And here I start 

 with the remark that, in comparing other spectra with the aurora, 

 it is, I think, too much the practice to trust to the coincidence 

 (more or less perfect) of one, or perhaps two lines out of many, 

 whereas we know by experience that most spectra have so well- 

 marked a general as well as special character, that when once 

 seen they are recognized afterwards with the greatest ease and 

 without measurements. Of course no two given spectra could 

 be considered identical unless their principal lines coincided ; but, 

 on the other hand, the coincidence of one or two lines out of 

 many without other features cannot be satisfactorily or conclu- 

 sively held to establish identity. 



In Professor Herschel's letter (Phil. Mag. No. 322, p. 71) 



o 



Professor Angstrom's representation of the "spectrum of the glow 

 discharge round the negative pole of air-vacuum tubes" is given, 

 in comparison with the aurora-lines and those of olefiant gas. 



I presume, from an examination of the spectra in this woodcut, 

 that those of the violet-pole and of olefiant gas have been acci- 

 dentally transposed. Assuming this to be so, yet the scale is 

 too small to work satisfactorily with, and I must fall back 

 upon my own spectrum -plates to found my remarks upon. I 

 should also state that (assuming the mistake above mentioned) 

 Professor Angstrom's projection and my own of the violet-pole 

 spectrum, still differ in some respects, though otherwise they 

 have a common resemblance. This will be seen on reducing 

 them to the same scale. 



As the general result of my observations and drawings, / see 

 no reason for giving to the violet-pole glow any special or distin- 

 guished place in comparison with the aurora, far less for assign- 

 ing to it the nearly absolute monopoly of the spectrum. It is 

 true that the line 7 in the violet-pole glow is in close coincidence 

 with one of the aurora-lines; but how are the brighter bands 

 a and /3 accounted for ? These, as I have before pointed out, 

 alone survive when the tube is placed at a distance; and it is true 

 they are thus reduced to shaded-off lines; but the difficulty still 

 remains, that they are conspicuous for their absence in the aurora- 

 spectrum ; and I cannot but think, on the whole, that if the violet- 



