PROFESSOR FORBES ON THE COLOURS OF THE ATMOSPHERE. 387 



2. Admitting that the blueness of the reflected light of the sky is an inherent 

 quality, of which we can give no account, we must next say that it is running 

 too fast to a solution to admit with Brandes that the red of evening is solely 

 caused by the colour of the air being complementary to its reflected tint. His 

 explanation of the variable redness of sunset, owing to the variable opacity of 

 white vapours allowing the redness to be more or less distinctly perceived, though 

 ingenious, is palpably wrong. The simplest experiments prove that the redness 

 is not merely apparent, but depends upon the admixture of the variable ingre- 

 dients of the atmosphere. The proof is the Prismatic Analysis of the sun's light, 

 and we may add, the observation of artificial lights in different states of the at- 

 mosphere, which at some times are seen in their natural condition, at others lose 

 all their rays but the red, and finally vanish in fogs with an intense red glare. 



3. If fogs and clouds modify the solar light on the principle of reflecting the 

 rays they do not transmit, why do not such fogs and clouds appear vividly blue 

 by reflected light, as Nollet supposed a foggy atmosphere must do to a spectator 

 placed beyond it ? 



4. If the vesicles constituting the clouds give to the colourless light falling upon 

 them the various hues of sunset, why, in the first place, do we not perceive bows 

 of various hues, as Kratzenstein did in operating on the small scale ; and how 

 comes it that clouds, identical in structure, nay the very same clouds, do not ex- 

 hibit sunset tints at any other time of day ? But the most convincing proof of 

 any, is simply to watch the progress of the solar rays tinging a cloud successively 

 with different hues, just as it would a lock of wool similarly placed ; or as it does 

 the snowy Alpine summits. Forster mentions an instance of detached cirro- 

 cumuli being of a fine golden-yellow, but in a single minute becoming deep red. 



5. To these unanswerable difficulties the prismatic analysis of the blue and 

 sunset tints of the sky superadds one conclusive against the theory of Newton 

 as it at present stands. The reflected blue and transmitted red-orange are not 

 colours of thin plates. They are derived from aU parts of the spectrum by the 

 mysterious process of transmission, which has preserved them and absorbed the 

 rest. It is hopeless at present to inquire what is the mechanical constitution of 

 the medium which has effected this alchemy. 



One question, however, which is quite within our reach, remains to be an- 

 swered. The colours of the sky cannot indeed be explained, if by explanation we 

 mean an ultimate analysis of the mechanism producing them ; but the theory of 

 absorption is incomplete until we can shew in what part of the course of the rays 

 of light, and under what varying circumstances, the different phenomena of colour 

 may be produced. Hassenfratz observed, that the light of the horizontal sun 

 was deficient, when analyzed by the prism, in all the violet and blue rays.* Sir 

 D. Brewster, making a similar observation with more care, has detected a speci- 



* Kamtz, Lehrbuch iii. 40. 

 VOL. XIV. PART II. 3 F 



