66 Prof. A. S. Herschel on the Spectrum of the Aurora. 



somewhat larger light clouds; and at about 1 or 2 inches of 

 barometric pressure a spark passes between the poles. This 

 spark is red ; it scarcely diminishes the strength of the 

 concomitant glow discharge; and it is far less luminous than 

 the white spark which begins to appear at 5 or 6 inches of pres- 

 sure, and may often be seen at first broken up along its length 

 into parts which are alternately white and red. The spectra, 

 like the general appearances, of these two forms of the spark are 

 quite distinct. 



I do not know if these several phases of the positive part of 

 the discharge have all been examined spectroscopically. They 

 pass into each other according to the shape and size of the tube 

 or flask, as well as the air-pressure; and it is difficult to say how 

 much of each is concerned in those observations which have been 

 made of air-spark spectra in comparisons with the aurora. No 

 one, so far as I know, has compared with it the negative-glow 

 spectrum so fully as Angstrom has now done; and it seems very 

 probable that its peculiar fitness for the comparison has been 

 overlooked — the feat of filling a bottle with the negative glow 

 discharge being certainly a novelty; if it is really true that he 

 succeeded in obliterating the positive brush entirely in its favour. 



The next remarkable novelty in the paper is the way in which 

 he proposes to account for the " citron " line of auroras ; for 

 there is evidently nothing of the kind in the negative glow, how- 

 ever well that answers to all the secondary facts of faint blue, 

 red, and greenish lines. If oxygen and its compounds are (as 

 has, I believe, been lately shown) strongly fluorescent, Tait 

 and Dewar have also proved, as shown by some of their expe- 

 riments this year, that they also possess powers of phosphorescence 

 — Geissler tubes shining for some time after the spark has passed 

 through them, from the production of ozone during the discharge. 

 When one of the globes of a phosphorescent " garland " tube 

 was heated over a Bunsen flame, that globe which was heated 

 did not shine after the spark had passed, apparently because, 

 as we know, a very little heat is sufficient to destroy ozone. 

 Whatever the way may be in which the ozone or otherwise elec- 

 trified gas remains self-luminous after the discharge, it seems 

 very Reasonable to suppose some action of the same kind (perhaps, 

 as Angstrom says, simply fluorescence) as common inj all 

 auroras, and that this produces the well-known auroral line. 



Pocket spectroscopes can, of course, do nothing further to fix 

 the position of the citron line; nor can they alone fix very ex- 

 actly the places of any of the fainter ones. But as every aurora 

 shows this strong monochromatic light, it might be used to bring 

 out a row of punctures transverse to the slit, as a divided scale in 

 the field of view whereby to map the fainter lines, or at any rate 



