530 PHILOSOPHICAL TRANSACTIONS. [ANNO 1784. 



off" so perpetually from the body of fire-balls, may possibly have some connection 

 with these streams. In the same manner the sound of explosions may perhaps 

 be brought to us quicker, than if it were propagated through the whole distance 

 by air alone. Should these ideas be well founded, the change of direction which 

 meteors seem at times to undergo, may possibly be influenced by the state of 

 the surface of the earth over which they are passing, and to which the streams 

 are supposed to reach. A similar cause may occasion the apparent explosion, 

 the opening of more channels giving new vent and motion to the electric fluid. 

 May not the deviation and explosion which appear to have taken place in the 

 tire-ball of the 18th of August over Lincolnshire, have been determined by its 

 approach towards the fens, and an attraction produced by that large body of 

 moisture ? 



3. A further argument for the electric origin of meteors is deduced from their 

 connection with the northern lights, and the resemblance they bear to these 

 electrical phenomena, as they are now almost universally allowed to be, in several 

 particulars. Instances are recorded, where northern lights have been seen to join 

 and form luminous balls, darting about with great velocity, and even leaving a 

 train behind like the common fire-balls. This train I take to be nothing but the 

 rare air left in such a highly electrified state as to be luminous; and some streams 

 of the northern lights are very much like it. The aurora borealis appears to 

 occupy as high, if not a higher, region above the surface of the earth, as may be 

 judged from the very distant countries to which it has been visible at the same 

 time; indeed the great accumulation of electric matter seems to lie beyond the 

 verge of our atmosphere, as estimated by the cessation of twilight. Also with 

 the northern lights a hissing noise is said to be heard in some very cold climates; 

 Gmelin speaks of it in the most pointed terms, as frequent and very loud in the 

 north eastern parts of Siberia; and other travellers have related similar facts. 



But, in my opinion, the most remarkable analogy of all, and that which tends 

 most to elucidate the origin of these meteors, is the direction of their course, 

 which seems, in the very large ones at least, to be constantly from or toward the 

 north or north-west quarter of the heavens, and indeed to approach very nearly 

 to the present magnetical meridian. This is particularly observable in those 

 meteors of late years whose tracks have been ascertained with most exactness; 

 as that of November 0.6, 1758, described by Sir John Pringle; that of July 17, 

 1771, treated of by M. Le Roy; and this of the 18th of last August. The 

 largest proportion of the other accounts of meteors confirm the same observa- 

 tion, even those of a more early period; nay I think some traces of it are per- 

 ceivable in the writings of the ancients. Whether their motion shall be from 

 the northern quarter of the heavens or toward it, seems nearly indifferent, as 

 the numbers of those going each way are not very unequal; I consider them in 



