1909-10.] Observations on some Spark-Gap Phenomena. 221 
by setting it at or near the negative pole. [This experiment, however, and 
most of the others to follow, will time and again seem to fail, owing to non- 
fulfilment of some one of the many limiting conditions. Thus the potential 
difference may not be near enough to the spark potential for free air, the 
dielectric may be too thick or too thin, or too broad, etc.] Even at the 
negative pole, however, it may be observed that if a dielectric plate be 
suddenly dropped vertically between the poles, a violent discharge at once 
takes place, and, curiously enough, more easily as a rule when the plate is 
so thin as to be easily ruptured by the discharge. In other words, the 
possibility of perforation seems to act as an extra inducement for the spark 
to pass, and it will not pass so freely if its path must lie round the edge of 
the plate. The whole phenomenon, and particularly the violence of the 
discharge, is most strongly marked when, at the instant of the plate’s fall, 
the terminals of the coil are giving off the characteristic brush discharge.* 
Again, if the dielectric (a sheet of cardboard, a thin plate of wax, glass, 
mica, etc.) be so set up that it is at right angles to the lines of the electrodes 
and can be moved freely in the direction from one electrode to the other, 
some interesting results may be noted. Thus, if each electrode is spherical, 
i.e. has, for example, a brass ball screwed on to its sparking end, motion of 
the card in the direction from the + electrode to the — electrode results in 
a great facilitation of the spark, while motion in the opposite direction 
distinctly acts as a deterrent. Thus, even in the case of a fairly wide gap 
(6-8 inches, say) with the dielectric set vertical in the middle a quick, sharp 
motion of quarter or half an inch towards the — pole will often induce a 
perforating discharge even though the potential difference is not high 
enough to produce a brush discharge in free air. And so a little paddle- 
wheel, whose paddles are made of the dielectric in use, may be set up so 
that the paddles when rotating pass, at the lowest point, between the 
electrodes from + to — and provoke a steady stream of sparks. The con- 
verse of this experiment — the “ quenching ” of an already sparking gap 
by rotation in the opposite sense — is more difficult to perform, but usually 
can be got to succeed with a little patience. The same results are to be 
observed when each electrode is fitted with a conical cap, or point. But if 
now one be spherical and one conical, an important difference comes in, for 
now the direction of motion for provocation of a spark is found to be always 
towards the conical electrode, no matter whether it be + or — . The 
dielectric itself, as will appear in § II., is in this case under a continuous 
and very sensible repulsive force tending to drive it towards the spherical 
electrode again quite independently of the polarity of the electrodes. 
* Of. Faraday, Experimental Researches in Electricity, series xii. 
