C 207 ] 
Upon the whole, this remarkable experiment 
feems to be made to the moll; advantage in the fol- 
lowing circumftances. Let the jar ftand upon the 
table ; let a thick brafs rod, infulated, Hand contigu- 
ous to the coating ; and, near the extremity of this 
rod, place the body that is to receive the explofion. 
This body muft be fix or feven feet in length, and, 
perhaps, fome inches in thicknefs, or be connected 
with a body of thofe dimenfions. Laftly, let the 
explofion be made with the difeharging rod refting 
upon the table, clofe to a chain, the extremity of 
which reaches within about an inch and a half of 
the coating cf the jar. In this cafe, the operator 
will hardly fail of getting a lateral explofion of an 
inch in length ; which fhall enter and leave the in- 
fulated body, without making any fenfible alteration 
in the electricity natural to it. 
With large jars, containing three or four fquare 
feet of coated glafs, bearing a very high charge, I 
make no doubt but that this experiment might be 
made to much more advantage ; but, at the time 
that I was engaged in this inveftigation, I happened 
not to have any fuch jar, and therefore only ufed 
one that contained half a fquare foot of coated 
glafs. 
If the interruption in the circuit, which is almofi: 
neceflary in thefe experiments, be made by intro- 
ducing a length of chain into it, rather than by 
making part of the explofion pafs along the tube, 
there is a medium in the length of chain, that an- 
fwers better than either a longer or a (horter circuit. 
In a long interrupted circuit, the eledric matter feems 
