418 M. Verdet on the Optical Properties developed in 



Hence, when a salt dissolves in water, the water and the salt 

 each contributes its own proper magneto-rotary power to the solu- 

 tion, and the rotation produced by the solution is the sum of the 

 separate rotations due to the molecules of the two substances. 



It is very probable that this law holds good with all solutions 

 of every kind, and with niLxtui'cs formed of liquids which have 

 no chemical action upon one another. But as yet I have not 

 made any experiments to confirm this. I have limited myself 

 to showing that the general bearings of the phsenomena are 

 the same; that, for instance, those salts whose aqueous solu- 

 tions have a high magneto-rotary power, behave in the same 

 manner when dissolved in alcohol or sether. This I have shown 

 in the case of several metallic chlorides which are soluble in these 

 liquids, namely, the bichlorides of mercury and of tin, and the 

 chlorides of cobalt and nickel, which latter two, though mag- 

 netic, behave like diamagnetic bodies. 



Most salts impart to their aqueous solutions magneto-rotary 

 powers superior to that of water. Sometimes, however, the 

 reverse is the case ; the salt contained in a given volume of the 

 solution exerting upon polarized light a less amount of action 

 than that effected by the water which it displaces. Thus the 

 solution of nitrate of ammonia given in Table I., consists of 43 

 parts of water and 57 of the nitrate : its density being 1*2566, 

 it follows that a unit volume contains 0"6660 of water and 

 0"5906 of the nitrate. The rotation being only 0'908ths of that 

 of distilled water, may be regarded as the sum of a rotation 

 0'666 effected by the water, and one of 0'243 produced by 

 the nitrate : the last number gives the small value 0"401 as the 

 molecular rotary power. Similar phsenomena would occur on 

 mixing water with a liquid whose magneto-rotary power is less 

 strong, such as alcohol, aether, or wood-spu'it. 



According to M. Bertin, solutions of the protosulphate of iron 

 appi'oach, in respect to their weakness in magneto-rotary power, 

 to solutions of nitrate of ammonia. M. Edmond Becquerel 

 made a similar observation with solutions of the protochloridc 

 of iron, and thought himself entitled to draw the general con- 

 clusion, that the rotation of the plane of polarization pi'oduced 

 by the influence of magnetism varies in inverse proportion to the 

 magnetic power of the body*. The experiments cited by M. 

 Edmond Becquerel himself do not admit of so absolute a con- 

 clusion. In fact, from his memoir we see that, if the rotation 

 of water be represented by 10, the rotation of two unequally 

 concentrated solutions of protochloridc of iron are represented 

 by 9 and 3, and that of a solution of sulphate of nickel by 13*55 ; 

 in other words, among the magnetic solutions examined, two 

 * See Annales de Chimie et de Physique, 3 ser. vol. xxviii. p. 334. 



