ell be employed with the production of glow be used, the
glow will appear at the place where the current leaves the ball, and that
will be the part directly opposite to the connection of the ball and rod
which supports it; but by increasing the tension elsewhere, so as to raise
it above the tension upon that spot, which can easily be effected
inductively, then the place of the glow and the direction of the current
will also change, and pass to that spot which for the time is most
favourable for their production (1591.).

1577. For instance, approaching the hand towards the ball will tend to
cause brush (1539.), but by increasing the supply of electricity the
condition of glow may be preserved; then on moving the hand about from side
to side the position of the glow will very evidently move with it.

1578. A point brought towards a glowing ball would at twelve or fourteen
inches distance make the glow break into brush, but when still nearer, glow
was reproduced, probably dependent upon the discharge of wind or air
passing from the point to the ball, and this glow was very obedient to the
motion of the point, following it in every direction.

1579. Even a current of wind could affect the place of the glow; for a
varnished glass tube being directed sideways towards the ball, air was
sometimes blown through it at the ball and sometimes not. In the former
case, the place of the glow was changed a little, as if it were blown away
by the current, and this is just the result which might have been
anticipated. All these effects illustrate beautifully the general causes
and relations, both of the glow and the current of air accompanying it
(1574.).

1580. Flame facilitates the production of a current in the dielectric
surrounding it. Thus, if a ball which would not occasion a current in the
air have a flame, whether large or small, formed on its surface, the
current is produced with the greatest ease; but not the least difficulty
can occur in comprehending the effective action of the flame in this case,
if its relation, as part of the surrounding dielectric, to the electrified
ball, be but for a moment considered (1375. 1380.).

1581. Conducting fluid terminations, instead of rigid points, illustrate in
a very beautiful manner the formation of the currents, with their effects
and influence in exalting the conditions under which they were commenced.
Let the rounded end of a brass rod, 0.3 of an inch or thereabouts in
diameter,