it seems to me impossible to resist the idea that it must be preceded
by a _state of tension_ in the fluid, and between the fluid and the zinc;
the _first consequence_ of the affinity of the zinc for the oxygen of the
water.

951. I have sought carefully for indications of a state of tension in the
electrolytic conductor; and conceiving that it might produce something like
structure, either before or during its discharge, I endeavoured to make
this evident by polarized light. A glass cell, seven inches long, one inch
and a half wide, and six inches deep, had two sets of platina electrodes
adapted to it, one set for the ends, and the other for the sides. Those for
the _sides_ were seven inches long by three inches high, and when in the
cell were separated by a little frame of wood covered with calico; so that
when made active by connexion with a battery upon any solution in the cell,
the bubbles of gas rising from them did not obscure the central parts of
the liquid.

952. A saturated solution of sulphate of soda was put into the cell, and
the electrodes connected with a battery of 150 pairs of 4-inch plates: the
current of electricity was conducted across the cell so freely, that the
discharge was as good as if a wire had been used. A ray of polarized light
was then transmitted through this solution, directly across the course of
the electric current, and examined by an analysing plate; but though it
penetrated seven inches of solution thus subject to the action of the
electricity, and though contact was sometimes made, sometimes broken, and
occasionally reversed during the observations, not the slightest trace of
action on the ray could be perceived.

953. The large electrodes were then removed, and others introduced which
fitted the _ends_ of the cell. In each a slit was cut, so as to allow the
light to pass. The course of the polarized ray was now parallel to the
current, or in the direction of its axis (517.); but still no effect, under
any circumstances of contact or disunion, could be perceived upon it.

954. A strong solution of nitrate of lead was employed instead of the
sulphate of soda, but no effects could be detected.

955. Thinking it possible that the discharge of the electric forces by the
successive decompositions and recompositions of the particles of the
electrolyte might neutralize and therefore destroy any effect which the
first state of tension could by possibility produce, I took a substance
w