390 PROFESSOR FORBES ON THE COLOURS OF THE ATMOSPHERE. 



temperature of the day and before sunset, the surface of the gi-ound, and likewise 

 the strata at different lieights in the atmosphere, begin to lose heat by radiation. 

 This is the cause of the deposition of dew, and consequently in severe weather 

 we have vast tracts of ah* containing moisture in that critical state which pre- 

 cedes condensation, and yet it may be exceedingly doubted whether any vapour 

 properly called vesicular is necessarily formed in this process. Be that as it may, 

 every accurate observer of nature in alpine countries wiU confirm me in stating, 

 that fine weather is almost invariably accompanied by the formation of dew on 

 exposed surfaces, and by the progressive depression of the moister strata, until at 

 length visible fogs are formed in the bottom of the valleys, and especially over 

 water.* This is the sm-est sign of a following fine day in mountainous regions. 

 Now Saussure in his ascent of Mont Blanc, " observed that the evening vapour 

 which tempered the sun's brightness, and half concealed the immense space he 

 had below him, formed the finest purple l)elt, encircling all the western horizon, 

 and as the vapom- descended and became more dense, became narrower and of 

 a deeper colour, and at last of a hhod-red."] Now this phenomenon corre- 

 sponds, I imagine, precisety to the development of colour which I have re- 

 marked in vapour in the act of being condensed, and De la Rive's remark, that 

 the nocturnal illumination of Mont Blanc takes place in serene evenings, when the 

 air is highly, charged with moisture, is to the same purpose. But a remark of 

 Mr FoRSTER, in his " Researches about Atmospheric Phenomena," J is even more 

 pointed, and is valuable, because his work is pre-eminently descriptive, rather 

 than theoretical. " Sometimes the tints in the twdhght haze come on so suddenly 

 and ai-e so ckcumscribed, as to induce a belief that very sudden and partial 

 changes take phxce in the atmosphere at eventide ; which may perhaps he some- 

 how connected with the formation of dew." He then records an observation made 

 2d November 1822. " Being about four o'clock in the evening, near Croydon in 

 Surrey, I observed a very beautiful western sky, caused by the bright edge and 

 dependent fringes of a light bed of cloud being finely gilded by the setting sun. 

 Some detached ciiTocumuli also, which formed the exterior boundaries of the 

 aforesaid cloud, were Mkewise of a fine golden-yellow, and the same colour ap- 

 peared in different clouds in other parts of the sky, while the scud-hke remains of 

 the nimbus floated along in the west wind below. In the course of about a quar- 

 ter of an hour, the lofty gilded clouds all assumed a deep red appearance, and' the 

 change was effected so suddenly, that while looking at them, I only took my eyes 



* For the reason why over water, see Davy's Paper, Phil. Trans. 1819. 



t Quoted by Harvby in Encyc. Metrop. Meteorology, p. *166. The cause of the purple light 

 mentioned here, probably arises from a mixture of the reflected blue of the pure sky (which is always 

 present when purple is seen) with the yellow-orange, which condensing vapour first transmits. I do 

 not think it at all necessary to aflirm, however, that pure air has no transmitted colour of its own. 



t Third edit. p. 87. 



