IN STUDYING QUESTIONS OF CHEMICAL MECHANICS. 315 
reciprocal proportion of the dispersion to the squares of the 
lengths of the fits is only an approximation of very general ap- 
plication, and not a property of the luminous principle, as had 
previously been supposed. This has been since completely 
proved by the observations made on the aqueous solutions of 
tartaric acid, since the dispersion of the planes of polarization 
is effected by it according to wholly different laws. But before 
these last phenomena were known, the dispersions observed in 
all other colourless active liquids produced successions of tints 
so approximately similar, when the analysing prism accomplished 
its revolution, that we might have thought them identical, and 
that we may still, without inconvenience, employ them as such 
in a multitude of observations. The exact restitution of the 
type ray to its primitive plane of polarization in the mixed 
‘system composed according to the theoretical proportion of vo- 
lumes, and the approximative return to the same plane, which 
is effected in all the other rays, demonstrate the molecular cha- 
racter of this kind of action by the most rigorous proof which 
can be conceived, since these restitutions are only effected by 
the final equivalence of an almost infinite multitude of deviations 
in contrary directions, experienced successively by the luminous 
elements during the whole length of their passage*. 
* I have repeated this experiment on two liquids whose mode of dispersion 
might have been presumed, and was found in fact much more similar. The 
first, which I shall designate by S, was syrup of pure cane-sugar diluted with 
water; the second, which I shall call S’, was another syrup of the same sugar, 
which M. Soubeiran had treated with hydrochloric acid, and then saturated 
with a great excess of very pure marble; so that it consisted of a solution of 
uncrystallizable sugar mixed with chloride of calcium. These two liquids were 
observed with the naked eye, in glass tubes of nearly equal length, taking for 
type of the deviations the extraordinary bluish violet tint. For S’ the length 
was 146™-25, and the deviation 25°33 \, directed toward the left of the 
observer. For S the length was 147™°75, and the deviation + 53° , di- 
rected toward the right; this reduced to 146™:25 by the law of proportionality, 
becomes 4.52°-46/'. I then mixed S! with S in the relations of volume 
inverse to those numbers, or as 207 to 100, and I observed the mixture through 
a tube 501™™ long, which more than tripled the effects of all the errors which 
I might have committed. The compensation was total. The extraordinary 
image was null in the azimuth of primitive polarization, and presented no sen- 
sible trace of coloration beyond that azimuth, just as if a completely inactive 
liquid had been interposed in the course of the ray. But I presume that the 
compensation is destroyed in time, either by the spontaneous fermentation of 
the liquid S, or by the action which the chloride may exercise upon it, which 
would cause a resultant toward the left. On this account I kept the mixture 
in Byte glass simply covered by a capsule, in communication with the air, 
and [ shall mention in the latter part of this memoir the effects which it will 
have undergone. 
