Rotatory Polarization of Liquids. 405 



unsatisfactory results, on account of the difficulty experienced, 

 with those which have a natural rotatory polarization, in deter- 

 mining the sensible tint. The essence of copaiba is more suitable 

 for experiment. I first determined the natural rotatory polari- 

 zation of a column of this liquid 10 centims. long, and found it 

 to be 23° 30'. The analyzer, placed so as to give the sensible 

 tint without the action of the magnet, marked 124° 30'. Mag- 

 netizing in one direction, I recovered the sensible tint at 118° 45'; 

 magnetizing again in the other direction, I recovered the same 

 tint at 130° 20' : this gives for the double of the polarization 

 11° 35', that of water being 8° 45' in the same circumstances. 

 We thus obtain 1*320 for the specific-polarization coefficient of 

 essence of copaiba. It may be remarked that the two magnetic 

 rotations give angles sensibly equal (5° 45' and 5° 50'), to the 

 left and to the right of 124° 30', the azimuth of the natural sen- 

 sible tint. It is not so with the essence of turpentine, the natural 

 rotatory polarization of which is greater. Starting from 145° 40 ; , 

 the angle corresponding to the natural sensible tint, we have, 

 according to the direction of magnetization, 141° and 151° 20 ; , 

 which gives 4° 40' in one direction and 5° 40' in the other, 

 together 10° 20' for the double of the angle of magneto-rotatory 

 polarization. Now, the corresponding angle for water being 

 8° 40', this gives 1*192 for the coefficient of the specific magneto- 

 rotatory polarization of the essence of turpentine. I do not 

 attach much importance to this number, on account of the diffi- 

 culty of procuring essence of turpentine always of the same com- 

 position. 



The same remark applies to the creosote of commerce, upon 



which I made several experiments, but will cite only one ; it gave 



2*259 for the coefficient of specific magneto-rotatory polarization. 



Creosote. Water. Ratio. 



19° 20' 9° 2*259 



We shall not at present discuss the results recorded in this 

 Section ; but we shall return to them at the end of this memoir, 

 when we shall have more data for the deduction of consequences 

 from them. 



§ 3. Influence of Temperature on the Magneto-rotatory Polariza- 

 tion of Liquids. 



According to some physicists, rise of temperature increases the 

 magneto-rotatory power of solid bodies (of glass, for example) ; 

 acording to others it diminishes that power. It is probable that 

 these divergences arise from the inequalities of the dilatation 

 presented by solids, according to the direction in which they are 

 observed — especially those which, like glass, have a structure 

 which is not uniform. For observations of this kind, liquids are 



