386 PROFESSOR FORBES ON THE COLOURS OF THE ATMOSPHERE. 



Probably uiDon the principle of multiplied reflections, the cases of preternatu- 

 rally protracted twilights may be explained, such as those recorded by Kamtz*. 



It is now time that we endeavour to sum up briefly the evidence we have 

 collected. 



If we exclude the theory of Leonardo da Vinci and Gothe, attributing the 

 colour of the sky to a mixture of light and shade ; and that of Muncke, which 

 would make it a mere optical deception, we shall find the chief principles which 

 have been maintained, reduced to three. 



(1.) That the colour of the sky is that reflected by pure air, and that all the 

 tints it displays are modifications of the reflected and transmitted light. This is 

 more or less completely the opinion of Mariotte, Bouguer, Euler, Leslie, and 

 Brandes. 



(2.) That the colours of the sky are explicable by floating vapours acting as 

 thin plates do in reflecting and transmitting complementary colours. This was 

 Newton's theory which has been adopted in whole or in part by many later 

 writers, and especially by Nobili. 



(3.) On the principle of opalescence and of specific absorption depending on the 

 nature and unknown constitution of floating particles. To this theory in its 

 various stages, we find Fabri, Melvill, Delaval, Count Maistre, and Sir D. 

 Brewster, attached. 



These different views are so easily blended, and have often been so far misun- 

 derstood even by their supporters, that it is impossible to draw any definite line 

 between them. I will notice a few of the leading points of difficulty which pre- 

 sent themselves to some of these opinions, and tend to restrict the field of 

 inquiry. 



1. The azure of the sky cannot, I think, with any probability, be referred to 

 the existence of those vesicular vapours which are supposed to act so important a 

 part in the mechanism of clouds. We have no evidence direct or indirect of their 

 existence, whenever the hygrometer is not affected, nor indeed where it does not 

 indicate absolute dampness. The atmosphere we know to be pre-eminently 

 transparent when loaded with uncondensed vapour. That vapour may be colour- 

 less, or it may not ; the presumption is, I think, that it has no colour, since the blue 

 of heaven is always most fully developed when the dryness of the air is intense ; 

 and that even at heights which render it in the last degree improbable that any 

 condensed vapour should exist at heights still greater. We are as ignorant of the 

 constitution of the parts of pure vapour, as we are of the parts of pure air : vesi- 

 cles are water, not vapour ; — to speak of films capable of reflecting definite colours 

 when no mater exists in the air, or the hygrometer does not indicate absolute 

 dampness, is to speak (as Berkeley said of Fluxions) of the ghosts of departed 

 quantities. 



* Lehrbuch der Meteorologie, iii. 38. 



