IN STUDYING QUESTIONS OF CHEMICAL MECHANICS. 333 
suitable aggregations of material elements individually inactive, 
the possibility of imitating them thus artificially would add much 
to its force. Now this is what Fresnel has effected by submit- 
ting a ray, polarized in a single direction, to the action of a thin 
lamina of gypsum, comprised between two rhomboids of glass, 
in each of which this ray undergoes internally a total reflexion ; 
the first before traversing the thin lamina, the second after 
having been modified by it. For on varying the constituent 
pieces of this apparatus, according to the conditions which the 
theory of undulations had suggested to him, he succeeded’ in 
impressing on the transmitted ray modifications exactly similar 
to those which an assigned thickness of essence of turpentine 
produces, not only for the direction of the deviation, and for the 
nature as well as the apparent order of succession of the tints in 
the different azimuths of the analysing prism, but also for the 
law of dispersion, in the inverse ratio of the squares of the lengths 
of their fits, which experiment had previously established as an 
almost exact approximation in all the active substances hitherto 
observed. It was then unknown that this identity of dispersion 
of the planes of polarization was not completely exact ; and above 
all, that tartaric acid dissolved in inactive liquids totally destroys 
it. But, if these remarks take from the apparatus of Fresnel the 
generality of theoretical application which he supposed it to pos- 
sess, as reproducing a relation of dispersion which would have 
been inherent in the nature of the luminous principle, the par- 
ticular imitation which he gives of that which is observed in the 
essence of turpentine, and consequently in rock crystal which 
follows a similar law, proves that there is no physical impossi- 
bility in the rotatory power of this crystal in the direction of its 
axis being an effect of aggregation, and not a molecular effect, as 
all the other circumstances related above agree in indicating. 
32. When a plate of rock crystal, perpendicular to the axis 
of the crystal, is traversed normally by a luminous ray polarized 
in a single direction, the power of double refraction, resulting 
from the uniform crystallization, is null on a ray thus circum- 
stanced, and cannot change the direction of its primitive polari- 
zation. Therefore, as such a change is then actually produced, 
_ it must necessarily result from an accessory cause, which, as we 
have just observed, resides, in all probability, in the mode of 
apposition of the crystalline laminz constituting each crystal. 
Whatever it may be, let us abstract it for a moment, and 
consider only the effects of the double refraction with a single 
VOL, IV. PART XIV. ZA 
