420 Researches on the Magnetic Rotatory Polarization of Liquids. 



is almost exactly the mean of the powers of the two liquids which 

 compose it — account being taken of the proportion of each in 

 the mixture, and of the alteration of density, in this case very 

 small. On the contrary, when the proportion of acid or of 

 water is more than two tenths, the rotatory power increases in a 

 much less proportion than the density, proving that the combi- 

 nation diminishes the molecular magneto-rotatory power. It is 

 true that in this case there is a strong chemical action, as is 

 proved by the great disengagement of heat. I am disposed to 

 believe that in all cases there is formed a compound of water and 

 acid, of which the magneto-rotatory power is always less than 

 the mean of the rotatory powers of the water and the acid in 

 combination, even allowing for the contraction, as we have seen 

 above. Only, when the proportions of the two substances mixed 

 are very different, the compound, dissolved in a large quantity 

 of the liquid which is in excess, cannot sensibly modify the mean 

 rotatory power. It is not so when the proportions of the two 

 liquids are nearly equal, the compound being in proportionally 

 much larger quantity in the mixture. Magnetic rotatory polari- 

 zation, therefore, might with advantage be used for the purpose 

 of distinguishing simple physical solutions or weak combinations 

 from definite chemical compounds. 



4. Our fourth conclusion is that the phenomenon of magneto- 

 rotatory polarization presents a means of penetrating into the 

 intimate constitution of bodies, and may thus be really service- 

 able to science. For example, the still very imperfect investi- 

 gation we have made shows already that the relations of ponde- 

 rable particles to the ether in which they are immersed do not 

 depend solely on the nature of the particles, but also on their 

 mode of grouping in the combinations which they form; for 

 simple mixture is not sufficient to modify these relations ; com- 

 bination is necessary in order that modification may take place. 

 I think, therefore, that it will be especially by operating on so- 

 lutions and isomeric bodies that we shall succeed in throwing 

 some light both on the nature of the phenomenon of magnetic 

 rotatory polarization and on the atomic constitution of bodies, 

 by determining for this purpose the differences existing between 

 the magneto-rotatory powers of isomers on the one hand, and 

 on the other the differences presented by the rotatory powers of 

 simple mixtures compared with those of true chemical combina- 

 tions. I shall endeavour, if I can procure the necessary sub- 

 stances, to study these two points more completely than I have 

 been able to study them in the present memoir. 



