84.) is made _negative_, it presents a
larger cone than when made positive; less of the fluid is thrown off, and
yet, when a ball is approached, sparks can hardly be obtained, so pointed
is the cone, and so free the discharge. A point held opposite to it did not
cause the retraction of the cone to such an extent as when it was positive.
All the effects are so different from those presented by the positive cone,
that I have no doubt such drops would present a very instructive method of
investigating the difference of positive and negative discharge in air and
other dielectrics (1480. 1501.).

1594. That I may not be misunderstood (1587.), I must observe here that I
do not consider the cones produced as the result _only_ of the current of
air or other insulating dielectric over their surface. When the drop is of
badly conducting matter, a part of the effect is due to the electrified
state of the particles, and this part constitutes almost the whole when the
matter is melted sealing-wax, oil of turpentine, and similar insulating
bodies (1588.). But even when the drop is of good conducting matter, as
water, solutions, or mercury, though the effect above spoken of will then
be insensible (1607.), still it is not the mere current of air or other
dielectric which produces all the change of form; for a part is due to
those attractive forces by which the charged drop, if free to move, would
travel along the line of strongest induction, and not being free to move,
has its form elongated until the _sum_ of the different forces tending to
produce this form is balanced by the cohesive attraction of the fluid. The
effect of the attractive forces are well shown when treacle, gum water, or
syrup is used; for the long threads which spin out, at the same time that
they form the axes of the currents of air, which may still be considered as
determined at their points, are like flexible conductors, and show by their
directions in what way the attractive forces draw them.

1595. When the phenomena of currents are observed in dense insulating
dielectrics, they present us with extraordinary degrees of mechanical
force. Thus, if a pint of well-rectified and filtered (1571.) oil of
turpentine be put into a glass vessel, and two wires be dipped into it in
different places, one leading to the electrical machine, and the other to
the discharging train, on working the machine the fluid will be thrown into
violent motion throughout its whole mass, whi