of the Atmosphere. 1 79 



proaches the meridian, the maximum polarization diminishes, 

 that it increases gradually as the sun descends from the meridian, 

 and reaches its maximum when the sun is very near the horizon, 

 the amplitude of this variation being about 0*09"*. 



This interesting result was obtained by adopting my measure 

 of the maximum polarization, namely, that when the sun's alti- 

 tude was 20°, the intensity of the point of maximum polarization 

 90° from the sun was equivalent to that which would be pro- 

 duced by reflexion from the surface of glass whose index of 

 refraction was 1*486 at an angle of 65° 80' f- From these data 

 the formulae of Fresnel give 0*64 as the measure of the intensity. 

 M. Bernard, taking the mean of two observations, found the 

 intensity to be 0*6523, differing only 0*012 from the measure 

 which I had obtained. 



In 1858, M. Liais, in crossing from France to Brazil, observed 

 the polarization of the atmosphere at the commencement of the 

 dawn {Vaurore) and at the end of twilight, in order to obtain a 

 measure of the height of the atmosphere. From observations 

 made at St. Domingo, and in the Bay of Bio Janeiro on the 1st, 

 2nd, and 3rd of December 1858, he found the height of the 

 atmosphere to be 340 kilometres f. During the same voyage, 

 M. Liais made the important observation that the zodiacal light 

 was not polarized §. 



In 1860, some interesting experiments on the polarization of 

 a beam of light diffused or dispersed by the smoke of different 

 substances were made by Professor M. G. Govi||. From their 

 relation to atmospheric polarization, they were repeated by Dr. 

 Bubenson, who extended them to the vapour of water. Professor 

 Govi found, as Dr. Bubenson did, that the diffused light was 

 polarized by reflexion or refraction, according as we viewed the 

 luminous cylinder perpendicular to its axis or in a direction 

 slightly inclined to it. The vapour of water, according to Dr. 

 Bubenson, produces the opposite effect. In 1846 I made ana- 

 logous experiments on the beam of light dispersed in passing 

 through certain fluids and solids ^[. 



In 1859, Dr. B. Bubenson, of the University of Upsal, under- 

 took a series of observations on the polarization of the atmo- 

 sphere. With the exception of a few made at Upsal in 1859, 

 they were carried on at Borne between the 6th of June and the 



* Comptes Rendus, vol. xxxix. p. 775, October 1854. 



f See Phil. Mag. S. 3. vol. xxxiv. p. 453, December 1847. 



X Comptes Rendus, vol. xliii. p. 109 (1854). 



§ Ibid. vol. xlvii. p. 450, September 1858. 



|| Ibid. vol. li. pp. 360 & 669. 



% Edinburgh Transactions, vol. xvi. p. Ill 



N 2 



