Causes of the Aurora Boreal is. 241 



by cold ; and if a preponderating quantity of hydrogen be added, it 

 is sufficient to produce its refractive poiver, for, according to New- 

 ton, " inflammable bodies refract much more light than the ratio of 

 their densities" " Hydrogen is inflammable, and remains perma- 

 nent over water, not combining with it in any notable degree."* Biot 

 ascertained that " hydrogen refracts light in a ratio independent of its 

 density higher than any other substance." It is much lighter than 

 common air, and consequently is capable of ascending in a gas- 

 eous form. It follows that any stratum or sheet of air, in which 

 more than a common portion of water is dissolved, and equally dif- 

 fused, accompanied by, or possessing an excess of hydrogen, be- 

 comes a powerfully refractive medium, with the capacity for reflec- 

 tion essential to the basis of the Northern Lights. What other for- 

 ces may be in exercise, or what agencies necessary to accumulate 

 this vapor, are not ascertained ; it may be analogous to, or perhaps 

 identical with the cause of the mirage, a well known refractive va- 

 por, which at various elevations, as well as on prairies and tracts of 

 desert, mocks the spectator with optical illusions, with visions of cit- 

 ies, villages, men, animals and sheets of water, spread in seeming 

 reality over the most arid sands ; while in mountainous countries, 

 such as the Hartz and the Scottish Highlands, fantastic images and 

 apparitions give rise to superstitious legends, and compel a belief 

 in supernatural revelations. 



I need not take time to detail the various opinions which have 

 been given of the constituent parts of this vapor, none of which are 

 conclusive ; but there is sufficient testimony that it exists in almost 

 every country, and that it rises to various elevations in the atmos- 

 phere.f 



* Henry's Chemistry. 



f It may be objected that the mirage always reflects objects horizontally to an ob- 

 server, who must look down at a certain angle, in order to perceive the im- 

 age produced by the refraction. But there is evidence that the same or a similar 

 vapor produces optical illusions at various elevations. Baron Humboldt saw stars 

 reflected, and moving like comets, with trains and sparkles after them, from the 

 Peak of Teneriffe, 10,800 feet above the level of the sea. They appeared 7° or 8° 

 above the horizon, and he at first suspected that they were the commencement of a 

 volcanic eruption from a mountain on an adjacent island ; but on further observation 

 with the telescope of a sextant, he perceived they were reflected images of those 

 stars then vertical to their apparent place. 



Another instance, shewing that this vapor does not invariably hover near the 

 earth's surface, occurred on the south side of Lon^ Island, in the summer of 1827 









