50 M. J.-L. Soret on Polarization 



242 sq. centims. inner coating, connected with the terminals of 

 the secondary coil. 



This discharge also lasts ^ second, and is similar to the 

 preceding, except that larger circles are made on the disk by the 

 dissipation of the carbon, and that there are fewer flashes, viz. 

 71. The total number of spark-holes in these flashes is 123. 

 Thus there are fewer flashes than in the experiments with the 

 platinum points, but the total number of spark-holes is the same 

 in each case. Hence there is, on an average, 1*34 spark to 

 each flash with the points, and 1*73 spark-hole to each flash 

 with the balls. 



Experiments have also been made with rotating disks formed 

 of " sensitized " paper ; and interesting results have been 

 obtained. 



VI. On Polarization by Diffusion of Light. 

 By J.-L. Soret *. 



On the Reflecting-power of Flames. 



I PUBLISHED, some months since, a first Notef, on the' 

 occasion of a memoir by M. G.-A. Hirn, in which he put 

 forth the hypothesis that the incandescent solid particles which, 

 according to Davy's theory, produce the brightness of flames, 

 become transparent at the high temperature to which they are 

 raised, and no longer possess any sensible reflecting-power. 

 One of the arguments which he advances in support of his hypo- 

 thesis consists in the fact that no phenomena of polarization are 

 observed in the light of a flame exposed to the rays of the sun. 



I indicated the results I had obtained by causing a pencil of 

 solar light to fall on lampblack, either when deposited on another 

 body, or at the moment of its formation — that is to say, when it is 

 in the state of smoke or of a smoky flame. In this latter case the 

 trace of the pencil of solar rays is perfectly visible : the part of the 

 flame which receives the raysappears bluish white, contrasting with 

 the reddish tint of the adjacent parts. If the trace be observed 

 with an analyzer, the light diffused in a direction at right angles 

 to the incident pencil is seen to be completely polarized in the 



* Translated from a separate impression, communicated by the Author, 

 from the Archives des Sciences of the Bibliotheque UniverseUe, July 1874, 

 pp. 1-26. An abstract of this memoir has been communicated to the Paris 

 Academy. 



t " On some Phenomena of Polarization by Diffusion of Light " 

 {Archives, Nov. 18, 1873, vol. xlviii. p. 231), Phil. Mag. S. 4. vol. xlvii. 

 p. 205. I take this opportunity to point out two important misprints in 

 that Note (Archives, pp. 235, 238): Phil. Mag. t. c. p. 208, line 12-13,/or 

 horizontal read vertical; p. 209, line 10 from bottom, for 50° read 90°. 



