From the accounts which have been collected of the polar lights, it would 

 seem that the phenomenon was less frequent in former ages than it is now ; but 

 it must be kept in mind that meteoric observations have not always been so 

 much attended to as at present. Aristotle describes the phenomenon with suf- 

 ficient accuracy in his book of meteors. Allusions are also made to it by 

 Pliny, Cicero, and Seneca ; so that it must have been witnessed by the an- 

 cients, even in the climates of Greece and Italy. The descriptions of armies 

 fighting in the air, and similar observations, in the dark ages, doubtless owed 

 their origin to the striking and fantastic appearances of the northern lights. 

 It is remarkable, however, that no mention is made by any English writer of 

 an aurora borealis having been observed in England from the year 1621 to 

 1707. Celsius says expressly that the oldest inhabitants of Upsala considered 

 the phenomenon a great rarity before 1716. In the month of March in that 

 year, a very splendid one appeared in England, and by reason of its brilliancy 

 attracted universal attentipn. It has been described by Dr. Halley in the Philo- 

 sophical Transactions, No. 347. Since then the meteor has been much more 

 common. A complete account of all the appearances of auroras recorded previ- 

 ous to 1754 may be found in the work of Mairan, " Traite de I'Aurore Boraele." 



The aurora is not confined to the northern hemisphere, similar appearances 

 being observed in high southern latitudes. An aurora was observed by Don 

 Antonio d'Ulloa at Cape Horn in 1745 ; one appeared at Cuzco, in 1744 ; and 

 another is described by Mr. Forster (who accompanied Captain Cook in his 

 last voyage round the world), which was seen by him in 1773, in latitude 58 

 south, and resembled entirely those of the northern hemisphere, excepting that 

 the light exhibited no tints, but was of a clear white. Similar testimony is 

 given by subsequent navigators. 



There is great difficulty in determining the exact height of the aurora bo- 

 realis above the earth, and accordingly the opinions given on this subject by 

 different observers are widely discordant. Mairan supposed the mean height 

 to be one hundred and seventy-five French leagues ; Bergman says four hun- 

 dred and sixty, and Euler several thousand miles. From the comparison of a 

 number of observations of an aurora that appeared in March, 1826, made at 

 different places in the north of England and south of Scotland, Dr. Dalton, in 

 a paper presented to the Royal Society, computed its height to be about one 

 hundred miles. But a calculation of this sort, in which it is of necessity sup- 

 posed that the meteor is seen in exactly the same place by the different ob- 

 servers, is subject to very great uncertainty. The observations of Dr. Rich- 

 ardson, Franklin, Hood, Parry, and others, seem to prove that the place of the 

 aurora is far within the limits of the atmosphere, and scarcely above the region 

 of the clouds ; in fact, as the diurnal rotation of the earth produces no change 

 in its apparent position, it must necessarily partake of that motion, and conse- 

 quently be regarded as an atmospherical phenomenon. 



L 



