66 ON THE DISTRIBUTION OF THE POLARISING FORCE 
ts 
attentive consideration. If the polarising forces depend solely 
on the mechanical condition of the particles of the glass, then 
it necessarily follows, that the central parts of the glass, which 
were in a state of variable expansion when it had a circular 
shape, are in a state of variable compression when the glass has 
received an elliptical form, and we are presented with a new 
law relative to the equilibrium of the cohesive forces in solids 
of variable density. But if the variation of density is merely 
the means of developing a new agent in the same manner as 
heat excites electricity in the tourmaline, or as pressure excites 
it in calcareous spar, then we cannot avoid regarding this agent 
in the same light as the electrical and magnetical fluid which 
are decomposed by certain mechanical operations, and distri- 
bute themselves according to regular laws. 
But whatever be the origin of the polarising forces, it be- 
comes a matter of great importance to discover the law of their 
distribution, when they are not controuled by opposite actions, 
and to apply this law to the explanation of the phenomena 
which they develope when they are either modified or extin- 
guished by the external form of the body in which they re- 
side. 
Those who have studied the papers to which I have already 
referred, cannot have failed to remark, that when the polari- 
sing forces are unconstrained in their developement, a negative 
structure is generated in the middle of the plate, when a posi- 
tive structure is generated at one or both of its edges ; and that 
the intensity of the negative, is to the intensity of either of the 
positive structures, as 10 to 16.02. 
In order to apply this principle to a circular plate, let KLHFG, 
Tig. 12. be a plate of this description, whose centre is O. Then 
if this plate were a part of the rectangular plate ABDC, mn, op 
would be the lines of no-polarisation which separate the internal 
negative 
