PKOFESSOR FORBES ON THE COLOURS OF THE ATMOSPHERE. 377 



first colour which vapours begin to reflect, it ought to be the colour of the finest 

 and most transparent skies in which vapours are not arrived to that grossness re- 

 quisite to reflect other colours, as we find it by experience."* In another propo- 

 sition, he says : " If we consider the various phenomena of the atmosphere, we 

 may observe, that when vapours are first raised, they hinder not the transparency 

 of the air, being divided into parts too smaU to cause any reflection in their super- 

 ficies. But when, in order to compose drops of rain, they begin to coalesce and 

 constitute globules of all intermediate sizes, those globules, when they become of 

 a convenient size to reflect some colours, and transmit others, may constitute 

 clouds of various colours, according to their sizes ; and I see not what can be ra- 

 tionally conceived in so transparent a substance as water for the production of 

 these colours, besides the various sizes of its fluid and globular parcels."! 



The theory of Newton, therefore, embraces the colour of clouds, whether by 

 reflected or transmitted light, as well as that of the blue sky. He applied a mo- 

 dification of the same theory to explain the coronce round the sun and moon. X 

 The air he seems to have believed to be devoid of colour, and the reflective parti- 

 cles to consist of vapour foreign to it. 



The idea of Mariotte of the inherent quality of the sky to reflect blue light, 

 was next prominently stated by Bouguer, who farther put it in so palpable a form 

 as to have been generally quoted since as a complete explanation of aerial colours. ^ 

 He observes, that as red light penetrates farther than blue (the reason is not men- 

 tioned), the latter is wholly reflected, whilst the former reaches the eye ; and this 

 theory was farther improved by later writers, by ascribing superior momentum to 

 the red rays, and inferior to the more refrangible ones. Smith, the author of the 

 System of Optics, states the same view, but with greater clearness. " The blue 

 colour of a clear sky," he says, "shews manifestly that the blue-making rays are 

 more copiously reflected from pure air than those of any other colour ; conse- 

 quently they are less copiously transmitted through it among the rest that come 

 from the sun, and so much the less as the tract of air through which they pass is 

 the longer. Hence the common colour of the sun and moon is whitest in the 

 meridian, and grows gradually more inclined to diluted yellow, orange, and red, 

 as they descend lower ; that is, as the rays are transmitted through a longer tract 

 of air ;" |j and so he explains the colour of the moon in eclipses by the altered 

 light refracted by the earth's atmosphere. 



Next, Euler (1762) maintained the same opinion as to the blueness of the 

 sky. " It is more probable," he says, " that all the particles of the air should 

 have a faintly bluish cast, but so very faint as to be imperceptible, until presented 



* Optics, Book ii. Part iii. Prop. vii. f Ibid. Prop, v. end. % Book ii. Part iv. Obs. 13. 

 § Traite d'Optique, p. 365-368. He likewise explains the coloured shadows noticed by Buffon. 

 II Smith's Optics, vol. ii. Remarks, 378. 



