266 THE ATIROUA BOREALIS. 



of some portion of the arch or corona, and upon the supposition of the 

 altitude and place being unchangeable for a very appreciable portion of 

 time. Two observers are, for instance, on the same meridian, and at 

 the same moment, notice the distance of the northern or southern edge 

 of the highest part of the arch, from some particular star. The difTer- 

 ence of this distance in degrees is the paralax, or angle which is sub- 

 tended by the arc of the meridian comprehended between the places of 

 the observers. This being known, the perpendicular height is easily 

 calculated by the rules of Plane Trigonometry. If now the correspond- 

 ing observations be accurately made, and at the same instant of time, 

 the paralax thus deduced must give the correct height. But it is safe to 

 say that no paralax hitherto has even approximated accuracy, except by 

 accident. It is amusing to read the accounts of the manner in which, 

 even scientific men have endeavored to persuade themselves that they 

 had demonstrated its altitude to be enormously great ; how out of the 

 notes of observers stationed at the same place they rejected all which 

 did not suit them ; and how they here allowed |° to 1°, and there 2° or 

 3° of probable error, and then announced that the aurora was far above 

 the limits of the atmosphere. Hence for the same arch or "auroral 

 cloud," we have an altitude of either 40 or 160 to 200 miles! It may 

 be doubted whether it be possible to obtain any thing like a reliable 

 paralax of so changeable a body as is an auroral streamer or arch, unless 

 it be very low, and hence the estimates of its altitude in high, agree far 

 better than those in low latitudes. 



It is, besides, a singular fact that the corona is always in or near the 

 elevated magnetic pole of each observer, and as no two observers can 

 have the same magnetic pole, this changing nearly with the latitude, 

 each sees a separate corona, as each sees a separate rainbow. However 

 similar or near absolute identity the coronas of two observers may ap- 

 pear, they are not the same, and so it may also be, to a certain extent, 

 with the arches, streamers, and other parts of the aurora. Hence it may 

 be just as impossible to obtain a correct paralax of a corona or streamer, 

 or even arch, as it would be to obtain one of a lunar or solar halo. — 

 The corona, at least, niust be an optical effect, depending upon some yet 

 unknoicn law of magnetism^ or magnetic condition of the particles of the 

 matter of the aurora, just as the rainbow and halo depend upon the 

 well known laws of light. All conclusions, therefore, concerning the 

 great altitude of the aurora, derived from a supposed paralax must a- 

 mount to notlyng more than approximate guesses; except that it is sat- 

 isfactorily determined that in middle latitudes it is considerably above 

 the region of the ordinary clouds. 



