Rotatory Polarization of Liquids, 397 



large enough to obtain this length, I had sometimes to content 

 myself with a shorter one, taking care, for the purpose of accu- 

 racy in the comparison, that in each case the length of the water 

 column should be the same as that of the liquid in question*. 

 I started from the principle that the ratio between the angles of 

 rotation is not affected by the absolute length of the liquid co- 

 lumns, provided this be the same in the case of the two liquids 

 compared. Moreover this principle results from that established 

 by M. Verdet in his beautiful researches on rotatory polarization, 

 in which he showed that the rotation is proportional to the 

 thickness of the substance submitted to the magnetic action, 

 provided that its intensity remain the same even when the thick- 

 ness varies. Nevertheless T wished to assure myself in a direct 

 manner that it actually is so; and I will here relate briefly the 

 experiments I made with this intention, which, while giving an 

 idea of the absolute magnitudes of the rotations I obtained f, 

 will at the same time pretty accurately indicate the maguetic 

 power of the apparatus I used. 



I filled a tube 30 centims. long, and 2 centims. in diameter, 

 successively with distilled water, rectified alcohol (sp. gr. 0*804), 

 sulphide of carbon, and iodide of ethyle, selecting these sub- 

 stances from among those which differ most in magnetic rotatory 

 power. 



I placed the tube between the two magnetic poles in such a 

 manner as to give the interpolar liquid column successively the 

 lengths of 25, 20, 15, 10, and 5 centims., by a corresponding 

 approximation of the polar surfaces — an approximation which 

 increased the intensity of the magnetic action, although the in- 

 crease did not quite compensate, by the magnitude of the rota- 

 tion, for the diminution induced by the shortening of the column. 



As the experiments on each liquid were made at different mo- 

 ments from the rest, and consequently with a different intensity 

 of the magnetizing current, only the results obtained from those 

 on each particular liquid are comparable ; but this is sufficient 

 for the end I had in view. The comparison of the liquids with 

 one another, which should give their specific magnetic rotatory 

 power, will be the object of the following section. 



For brevity's sake, I limit myself to giving, for each of the 

 four liquids submitted to experiment, the angles of rotation ob- 

 served when the lengths of the column were 25 centims. and 5 



* I ought to mention that the diameter of the column exercises no in- 

 fluence on the absolute magnitude of the rotation ; this is influenced by 

 the length alone. I gave the experimental proof of this result in the memoir 

 before mentioned. 



t The pile was composed of only 40 pairs ; with a stronger pile I ob- 

 tained, as will be seen, in other experiments, rotations of greater magnitude. 



