[ 460 ] 
For when a quantity of the eleCtric fluid is forced into 
the apparatus , which fupports the two balls, we fhould 
from its elaftic principle, expeCt it to pafs out again 
immediately: whereas the faCt is, that it paffesoutby 
flow degrees ; and takes a confiderable time in eva- 
cuating the apparatus effectually. Some power there- 
fore muft hinder the fluid, at leaft, in fome meafure, 
from efcaping : and that power mull be exerted at, or 
near, the furface of the body. 
To fay it is detained by an attraction of the body, 
will not anfwer the purpofe : for the power which is 
fuppofed to draw the fluid into it, muft certainly be 
fufficient to hinder it from pafling out. Now by the 
experiment, the fluid does pafs out, though flowly : 
this power therefore, which reflfts its pafling out, can 
be no other than what arifes from the medium we have 
proved to be fpread upon the furfaces of bodies. 
The evidence in favour of this doCtrine is greatly 
flrengthened by the following experiments, the three 
firft of which, are well known to eleCtric en- 
quirers. 
When glafs is properly electrified, and held over 
the wooden part of the apparatus (c) at the diftance 
of fix or eight inches, and there continued for a time, 
the balls are feparated to a confiderable diftance. 
But upon taking away the glafs, the feparation is 
at an end, and there are no eleCtric figns remaining 
in the balls. 
Thefe appearances therefore argue, that no part of 
the eleCtric fluid, appertaining to the excited glafs, 
pafled from it into the wood. And that the caufe, 
which ohftruCted its paflage, is a refinance , exerted at 
or near, the furface of the wood : becaufe we know, 
from a variety of experiments, that the repujfive power 
5 of 
