534 Progress in Science. [October, 
find that it diminished, and on further approach became zero, while it increased 
on slightly removing the rod. If the rod was withdrawn very slowly the di- 
vergence remained zero, but rose again to the former amount as the rod was 
further withdrawn. Frequent repetition of the experiment gave similar 
results, only, he remarks, the electrical source must be a pretty powerful one. 
After various futile attempts to explain the matter, there arose the doubt 
whether the divergence which appears after making a rubbed caoutchouc rod 
touch the ball of the electroscope was really due to negative electricity (—E). 
He became convinced that while the rod is, of course, negatively electrified, 
yet the leaves diverge with positive electricity (+E). He thus describes 
his experiments :—‘‘ Let a caoutchouc rod be rubbed with cat-skin and 
brought near the knob of a Fechner ele&troscope” (in which, it may be 
explained, there is only one leaf suspended between the opposite poles of a 
pile). ‘The leaf then moves to the + pole. The rod is therefore — eledtri- 
fied. Nowrub the rod again, and touch with it the knob of the gold-leaf 
electroscope’”’ (with two leaves). ‘‘Then remove the rod. Next, let the knob 
of this charged electroscope be brought near the knob of a Fechner electro- 
scope (uncharged). The leaf of the latter moves to the — pole; the leaves 
of the former, therefore, diverged with +E; and thus the electroscope became 
+ electrified through contact with the electrified rod. It remained to ascer- 
tain the reason of this fact, which, once ascertained as a fad, removed the 
difficulty in the first experiment. When the strongly — electrified rod is 
brought near the knob, there takes place decomposition of the eletricity in the 
electroscope. The +E flows into the knob, in which it is retained by the rod ; 
the —E flows into the leaves, which diverge in consequence. While thus in- 
fluenced by the rod —E escapes, and there is a progressive collection and re- 
tention of +E in the knob. At the moment of contaé& between knob and 
rod, the rod communicates the —E present at its point of contac to the knob, 
and this neutralises in the latter a corresponding quantity of +E. As, how- 
ever, the excited parts of the rod not in immediate contact with the knob do 
not give up their —E, this surplus of —E still carries on the decomposition 
and retention of electricity in the electroscope; so that there comes to be 
much more retained +E in the knob than —E in the leaves (for a constant 
waste of —E takes place in the electroscope). If the rod is now slowly re- 
moved its retentive influence on the knob decreases, and a certain quantity of 
—E flows into the leaves, neutralising there a certain quantity of —E. When 
the rod is just so far removed that as much +E can flow from the knob into 
the leaves as there is —E in these, the leaves become unelectrified and the 
divergence zero, On further withdrawal, still more of the hitherto retained 
+E flows from the knob to the leaves, and now commences a divergence with 
+E. When the rod is entirely removed, the whole of the hitherto retained 
+E is free, and produces strong divergence. As the rod is approached again, 
these phenomena are witnessed in reverse order. This theory rests on the 
supposition that the +E which has been separated and retained by the rod 
preponderates over the —E communicated to the electroscope by contact 
(which is easily understood, as an electrified non-condudctor gives up its elec- 
tricity only at the point of contact). For this preponderance to take place it 
is necessary, as has been said, that the electric source should be a very active 
one. It can, further, be easily shown that in such circumstances the —E 
flows away from the electroscope (without conduction) if the electrified rod be 
brought near the knob without touching it. In this case no ele¢tricity can 
flow dire@ly from the rod to the knob, and yet, if after a few seconds the rod 
be withdrawn, the leaves diverge considerably with +E. The explanation is 
simple. The fa&, however, that it is possible, by contact of the electroscope 
with a negatively electrified body, to obtain + divergence, appears to me one 
of considerable importance. Thus, supposing, in the absence of a pile elec- 
troscope, it were required to determine whether a body rubbed with a certain 
kind of material became + or — eleétrified, it would be a very likely method 
to first cause divergence in the ele@troscope leaves with a rubbed caoutchouc 
rod, and then test the electric state of the body in question by its influence on 
the divergent leaves, and call it + or — accordingly. On the former suppo- 
