EXTRACTS FROM A LETTER, &C. 41 



Edit. London. This curious and beautiful phenomenon may- 

 be seen every line summer's evening in this and perhaps in all 

 other countries, where serenity is united to a cloudless sky. It 

 is caused by the prismatic effect of the atmosphere upon the 

 sun's departing rays. Soon after sun-set a belt of a yellowish 

 orange colour is seen to extend itself along the eastern horizon, 

 this belt ascends in the same proportion as the sun descends, 

 being about one degree in breadth ; in contact with the first, 

 appears a second belt below, of a dark blue colour, and about 

 the same breadth as the first, both belts being tolerably well de- 

 fined and of a uniform colour throughout: when the double 

 belt has risen a little above the horison, the azure sky may be 

 seen below, and as the belts continue to ascend they become 

 fainter, until at length the prismatic rays meeting with no va- 

 pors sufficiently dense to reflect their colours, the whole pheno- 

 menon dissolves into pale celestial light; the belts disappear at 

 about 6 or 7°. of altitude. This phenomenon merits some 

 attention; it exhibits as upon a skreen that species of light, 

 which after a greater angular dispersion, arriving at the moon's 

 orbit, faintly illumines her disk during the time of a total 

 eclipse. 



It would seem to result from the above appearances, that if 

 a prism were formed of atmospheric air, the solar ray would 

 be separated thereby into two colours only, a yellow orange, 

 and a blue : it is known to opticians that the compound colour 

 of orange and yellow, and the colour which Newton calls indi- 

 go, comprise within themselves the seven primitive colours, that 

 is, united they ought to form white; we ought not therefore 

 to reject this effect of atmospheric air, because dissimilar to the 

 prismatic powers of such diaphanous bodies as are best known 

 to us; modern experiments have shewn that refracting bodies 

 possess very different dispersive powers ; and when we reflect 

 upon the heterogeneous nature of our atmosphere, composed 

 of at least three permanently elastic fluids, with the adventi- 

 tious mixture of perhaps a hundred others, subject from che- 

 mical affinity to perpetual resolution and composition, disolving 

 at all times a great proportion of aqueous fluid, and the whole 

 pervaded by the electric fluid; shall we presume to doubt, 

 that nature has it in her power to compose a refracting body, 



H 



