

on the Optical Phenomena of the Atmosphere. 15 



steeper than usual. In one single instance the transparency of 

 the atmosphere seems to increase the size of objects, and that is 

 when the chain of the Alps are observed from the plain to the 

 south or north. In moist weather, generally before the descent of 

 rain, the mountains appear darker and at the same time some- 

 what larger. This illusion appears to be due to the fact, that in 

 the latter case they are much more clearly and sharply defined 

 against the horizon*. 



The transparency of the atmosphere has the power to modify 

 in a great measure the magnitude of the prospect commanded 

 by a great altitude. This is never so great as to permit of being 

 calculated from the curvature of the earth and the refraction, 

 for in the lower portions the prospect is always considerably 

 limited by vapoursf. This explains why we see more clearly 

 looking from below upwards, than from a height downwards. 

 In the latter case, however, another influence operates. The 

 objects seen from above exhibit a uniform obscure colouring, 

 and do not present the same striking contrasts among themselves 

 as the rocks and the snow-covered mountains against the sky. 

 The greater transparency of the upper regions of the atmosphere 

 is strikingly exhibited when we direct our glance to higher 

 summits. It is surprising how plainly the latter stand out before 

 us, and with what distinctness we can recognise the objects which 

 rest upon them. The reason of this is, that we look through a 

 higher and more rarified atmospheric region. 



The intensity of the rays of light can also be approximately 

 determined by their chemical action upon colours J ; although the 

 results depend upon the material, &c, the increase of intensity with 

 the height is plainly manifested. We made use of strips of paper 

 on which a uniform wash of carmine was laid. In each experi- 

 ment one-half of each strip was exposed to the sim from 11 

 to 2 o'clock, while the other half was shaded by an opake screen. 

 The altered colours were imitated by carefully mixing together 

 carmine and white in different proportions (the exact process 

 may be learned where the cyanometer is described). In this 

 way we obtained the following corresponding quantities ; the 



* A very simple practical rule to calculate the circle of view from the 

 height is that used by seamen. The square root of the height in Hamburg 

 feet" gives the radius in sea miles, 60 to the degree. The refraction is here 

 taken into account. A Hamburg foot is =0-^(i met. = 127'0 Parisian 

 lines. For heights of 12,000 feet, we obtain US sea miles = 29 geogra- 

 phical miles; for Mont Blanc, Saussure gives 136' sea miles {Voyages, 

 vol. iv. ito, p. 194). 



f Tin- tame iraa observed by Humboldt and Bonplaud (Tubleau Physique 

 des Regions Equmoxiales. Paris, 1807, p. 135) ; and also by Gay-Lussac 

 during his neiia! j ui ii 



; Saussure, iWm. de Turin, vol. iv. p. 441-453 ; Voyages, vol. iv. p. 297- 



