On Material Molecules and the Etlierial Medium. 251 
an elliptic cylinder the same as a biaxal crystal, these phe- 
nomena ceasing as soon as the glass becomes uniformly 
heated. If a cylinder of glass, which has been heated to 
redness, be rapidly cooled at the exterior, the glass becomes 
permanently doubly-refracting, the position of the molecules 
being in this case permanently altered. 
These, and the phenomena produced in crystals by the 
combined effects of heat and pressure, may all be accounted 
for by the different changes in different directions, which 
take place in the position of the material molecules by the 
action of these causes. 
The next question is, whether the optical properties of 
crystals are determinated solely by the mass of the mole- 
cules and their arrangement in space, or are they dependent 
on the nature of the molecule, that is, on its chemical con- 
stitution, or the kind of matter of which it is built up ? 
The principal facts bearing on this point, which at first 
sight appear to show that such is the case, were observed 
by De Senarmont, who found that a change in the compo- 
sition of the molecule, produced by isomorphous replace- 
ment, alters the angle between the optic axes. But the 
change thus produced may, in reality, be due to the altera- 
tion in the arrangement, and to the change of mass of the 
molecules, consequent on isomorphous replacement. It may 
be objected, that such a supposition is contrary to the very 
meaning of isomorphism, which asserts that isomorphous 
substances can replace each other in all proportions, without 
any alteration in the form of the crystal, that is, in the 
arrangement of the molecules. It has, however, never been 
established, that the angles between the faces of crystals 
remain unaltered by isomorphous replacement ; on the con- 
trary, it appears to be very probable that this is not the case, 
from the very fact that isomorphous substances differ, in 
some cases considerably, in their " angular elements," as 
determined by the angles between the faces of crystals. 
In reality, there is no such thing as absolute isomorphism. 
In the well-known case, for instance, of the isomorphism of 
the carbonates of calcium, lead, strontium, and barium, in 
the minerals aragonite, cerussite, strontianite, and witherite, 
VOL. IH, 2 K 
