TIIE SKY OF Tnti ALPS. 269 



mountain sprang forth with astonishing solidity and detach- 

 ment from the surrounding air. The changes of the Dom 

 were still more wonderful. A vast amount of light could 

 be removed from the sky behind it, for it occupied the po- 

 sition of maximum polarization. By a little practice with 

 the Nicol it was easy to render the extinction of the light, 

 or its restoration, almost instantaneous. When the sky was 

 quenched, the four minor peaks and buttresses, and the 

 summit of the Dom, together with the shoulder of the Al- 

 phubel, glowed as if set suddenly on fire. This was imme- 

 diately dimmed by turning the Nicol through an angle of 

 90°. It was not the stoppage of the light of the sky behind 

 the mountains alone which produced this startling effect ; 

 the air between them and me was highly opalescent, and 

 the quenching of this intermediate glare augmented re- 

 markably the distinctness of the mountains. 



On the morning of August 24th similar effects were fine- 

 ly shown. At 10 A. m. all three mountains, the Dom, the 

 Matterhorn, and the Weisshorn, were powerfully affected 

 by the Nicol. But in this instance also the line drawn to 

 the Dom being accurately perpendicular to the direction of 

 the solar shadows, and consequently very nearly perpen- 

 dicular to the solar beams, the effects on this mountain were 

 most striking. The gray summit of the Matterhorn at the 

 same time could scarcely be distinguished from the opales- 

 cent haze around it ; but when the Nicol quenched the 

 haze, the summit became instantly isolated, and stood out 

 in bold definition. It is to be remembered that in the pro- 

 duction of these effects the only things changed are the 

 sky behind and the luminous haze in front of the moun- 

 tains ; that these are changed because the light emitted 

 from the sky and from the haze is plane polarized light, 1 

 and that the light from the snows and from the mountains 

 being sensibly unpolarized, is not directly affected by the 



1 Defined at page 255. 



