﻿of the Sky, and on the Polarization of Light. 393 



twists the plane of polarization 90°. On these and kindred points 

 experiments are still in progress*. 



The idea that the colour of the sky is due to the action of finely 

 divided matter, rendering the atmosphere a turbid medium through 

 which we look at the darkness of space, dates as far back as Leonardo 

 da Vinci. Newton conceived the colour to be due to exceedingly 

 small water particles acting as thin plates. Goethe's experiments in 

 connexion with this subject are well known and exceedingly instruc- 

 tive. One very striking observation of Goethe's referred to what is 

 technically called "chill" by painters, which is due no doubt to 

 extremely fine varnish particles interposed between the eye and a dark 

 background. Clausius, in two very able memoirs, endeavoured to 

 connect the colours of the sky with suspended water vesicles, and to 

 show that the important observations of Forbes on condensing steam 

 could also be thus accounted for. Briicke's experiments on preci- 

 pitated mastic were referred to in my last abstract. Helmholtz has 

 ascribed the blueness of the eyes to the action of suspended particles. 

 In an article written nearly nine years ago by myself, the colours of 

 the peat-smoke of the cabins of Killarneyf and the colours of the 

 sky were referred to one and the same cause, while a chapter of the 

 * Glaciers of the Alps,' published in 1860, is also devoted to this 

 question. Roscoe, in connexion with his truly beautiful experiments 

 on the photographic power of sky-light, has also given various in- 

 stances of the production of colour by suspended particles. In the 

 foregoing experiments the azure was produced in air 3 and exhibited a 

 depth and purity far surpassing anything that I have ever seen in 

 mote-filled liquids. Its polarization, moreover, was pe? feet. 



In his experiments on fluorescence Professor Stokes had conti- 

 nually to separate the light reflected from the motes suspended in 

 his liquids (the action of which he named "false dispersion") from 

 the fluorescent light of the same liquids (which he ascribed to " true 

 dispersion "). In fact it is hardly possible to obtain a liquid without 

 motes, which polarize by reflection the light falling upon them, truly 

 dispersed light being unpolarized. At p. 530 of his celebrated me- 

 moir " On the Change of the Refrangibility of Light," Prof. Stokes 

 adduces some significant facts, and makes some noteworthy remarks, 

 which bear upon our present subject. He notices more particularly 

 a specimen of plate glass which, seen by reflected light, exhibited a 

 blue which was exceedingly like an effect of fluorescence, but which, 

 when properly examined, was found to be an instance of false dis- 

 persion. " It often struck me," he writes, "while engaged in these 

 observations, that when the beam had a continuous appearance the 

 polarization was more nearly perfect than when it was sparkling, so 



* Sir John Herschel has suggested to me that this change of the polarization 

 from positive to negative may indicate a change from polarization by reflection to 

 polarization by refraction. This thought repeatedly occurred to me while looking 

 at the effects ; but it will require much following up before it emerges into 

 clearness. 



t 1 have sometimes quenched almost completely, by a Nicol, the light dis- 

 charged normally from burning leaves in Hyde Park. The blue smoke from the 

 ignited end of a cigar polarizes also, but not perfectly. 



