THE AURORA. 471 



auroral display; but some of the brightest have been seen at full 

 moon. Some have said that auroras could be seen in daylight, but 

 this sounds like a ghost story. 



It has been almost impossible to photograph the aurora, although 

 in the winter of 1882-83 a Swede at Spitzbergen claimed to have 

 secured a faint image in eight minutes and a half. This was done 

 before access was had to the orthochromatic plates, color screens, or 

 lightning emulsions of to-day. 



Some observers claim that the stars scintillate less when seen 

 through an aurora ; but this is caused, according to Montigny, by the 

 presence of fog. On the contrary, it is clearly proved that the scin- 

 tillation increases during any magnetic disturbance, even when the 

 latter is not accompanied by an aurora. 



As to the study of the nature of the light of an aurora there are 

 two methods: by the polariscope and by the spectroscope. By the 

 former it is easy to recognize whether it is a natural light emanating 

 from a self-luminous body, or whether it reaches the eye after under- 

 going one or more refractions or reflections. Biot, in 1817, in the 

 Shetlands, could not discover the smallest trace of polarization; this 

 has been confirmed by Macquorn, Rankine, and ITordenskiold, and 

 proves that the light of the aurora is not, like that of rainbows and 

 halos, the result of reflection or refraction, but is itself luminous. 



This important discovery is confirmed by the spectroscope. If 

 the light emanating from a solid or liquid incandescent body is passed 

 through the spectroscope, the resulting spectrum is continuous; if, on 

 the contrary, the source of the light is gaseous, the spectrum is com- 

 posed of a certain number of bright lines or stripes separated from 

 each other by dark intervals. The number, the position, and the 

 brilliancy of these bright lines depend upon the chemical constitution 

 of the glowing gaseous body. The spectrum of the aurora was 

 studied by Angstroem, in 1866, for the first time, and is essentially 

 a spectrum of lines, hence gaseous by nature; it can not, therefore, 

 be due to a reflection of the light of the sun, as has been supposed. 

 The spectrum of the aurora runs the gamut from red to yellow, green, 

 blue, and even violet, the last line of the spectrum having been seen 

 but once, by Lemstroem in Lapland. Some of the lines are very simi- 

 lar in position to the spectrum of the electric spark or of lightning. 

 The fourth line has not been found in any known body, and Ang- 

 stroem attributes it to phosphorescence or fluorescence. For in- 

 stance, oxygen is phosphorescent, and there is an abundance of ozone 

 created by the aurora. A drop of sulphate of quinine has been made 

 luminous by the action of the rays of the aurora, and so also has 

 the double cyanide of platinum and potassium. 



But, although we are getting nearer to the truth as to the nature 



