336 Mr. J. Rand Capron on the Auroral 



being so numerous that a coincidence of some of these with 

 the coarser aurora lines may after all be purely accidental. 

 (In ' Photographed Spectra' I have given the spectrum of a 

 meteorite when burnt in the electric arc, which presents several 

 hundred lines in the portion of the spectrum lying between 

 F and H alone.) 



In regard to Prof. Groneman's height of aurorae as com- 

 pared with Prof. Herschel's deductions, it may be desirable to 

 mention two recent authorities on the subject as giving heights 

 very close upon those assigned by Prof. Herschel : viz. Herr 

 Sophus Tromholt, in an article on Aurorae (' Nature,' vol. xxvii. 

 p. 295), speaks of an opinion he has formed as to the height 

 of the aurora, viz. 150 kilometres (90 miles) ; Baron Nor- 

 denskjold ('Nature,' vol. xxv. pp. 319-321), in the " Scientific 

 Work of the Vega Expedition," Part I. pp. 401-452, speaks of 

 our globe as being adorned with an auroral crown whose inner 

 edge was usually, during the winter of 1878-79, at a height 

 of about 0*03 radius of the earth above its surface (that is, 

 about 115 miles). 



While on this part of the subject, it may be useful to refer 

 to the vacuum experiments of Drs. De La Rue and Miiller, 

 detailed by them in a paper on the height of the Aurora 

 Borealis read before the Royal Society ^ Nature,' vol. xxii. 

 p. 33). Appended to this is a table of deductions from actual 

 observations, showing, amongst other particulars, a scale of 

 miles of auroral heights, and remarks on the character of the 

 electric glow at such heights. 11*58 miles gave a full red 

 glow, 27*42 a carmine, 32*87 a salmon-coloured, 37*67 maxi- 

 mum brilliancy, 81*47 pale and faint, and at 124*15 miles " no 

 discharge could occur." A note appended to the table states 

 " it is conceivable that the aurora may occur at times at an 

 altitude of a few thousand feet"*. 



The experiments, however, are open to the remark that the 

 discharge was taken in hydrogen gas and not in air, and that 

 there was an absence of any observation of aurora-like lines or 

 spectrum in connexion with the discharge. Otherwise the 

 experiments would go to show that the auroral discharge 

 generally takes place at a lower level than most observers 

 have ascribed to it, and, if they could be depended on as 

 fulfilling all the conditions of the aurora, would place the beam 

 at a height of 133 miles within the region where no discharge 

 could take place. Possibly, however, the medium in which the 

 discharge was taken might affect the results obtained to the 

 extent of this discrepancy. 



* Dr. De La Rue acquaints me he lias further experiments in progress.— 

 J. R. C. 



