328 BIOT ON THE EMPLOYMENT OF POLARIZED LIGHT 
28. In conclusion, of the five presumptive kinds of physical con- 
stitution which the systems above considered may present, there 
are four the realization of which may be detected by optical laws 
observable experimentally. The only one which eludes these 
determinations is that in which the inactive liquid, being-insuf- 
ficient in quantity to saturate completely the weight P of the 
active substance present, would distribute itself equally among 
all the molecules of that substance, so as to form with it chemi- 
cal groups imperfectly saturated and identical among themselves, 
but whose proportions would vary progressively according to the 
different values of the relation 2 The existence of this excep- 
tional case will be manifested by the special forms of variabi- 
lity, and exclusive of the four others, which the characteristic 
. a(P+ 4K) 
function = pBp 
forms above enumerated include all the cases of constitution 
of the mixed systems chemically supposable. If others were to 
be imagined, the optical effects which they should produce may 
be in like manner characterized theoretically, according to the 
definitions which we should assign to their state, and the ob- 
servation of the effects really produced would show if the sup- 
posed case of constitution were realized or not. 
29. All the preceding conclusions, as well as the reasonings 
from which they are derived, suppose essentially that the devi- 
ations impressed on the planes of polarization of the luminous 
rays, under the circumstances in which they are applied, are 
solely produced by a molecular action proper to the constituent 
groups of the media observed, considered as being in an abso- 
lute state of disaggregation; so that the polarized ray be solely 
influenced by the power of deviation which they successively 
exercise upon it, independently of one another. It becomes 
then indispensable to specify the physical tests by which we 
may ascertain that this condition of independence is or is not 
fulfilled, or only incompletely. For these three cases occur in 
the experiments, and we should arrive at very erroneous re- 
sults if we did not distinguish them before applying the formulze 
above established. I proceed, therefore, to assign them special 
characters in the following paragraphs. 
30. To make a legitimate use of the formule established in 
the foregoing paragraphs, we must be assured that the deviations 
will present; at least, admitting that the five 
