t *9^ ] 
tion of the uniform communication of electricity, 
either that of the infide or that of the outlide of the 
jar ; and, together with the former experiment?, 
convinced me, that this lateral fpark by no means pro- 
duced the effeCt that might have been expected in 
communicating electricity. But, with the next fet of 
experiments, the difficulty began a little to clear up, 
and continued to do fo gradually, till I gained all 
the fatisfaCtion I could wifh for, with refpeCt to this 
puzzling phenomenon. 
The firfl time that I was able to vary the electricity 
of the infulated body placed near the eleCtric circuit, 
or of the bodies that formed the circuit (which I now 
began to attend to), by any different adjuftment of 
the apparatus, was on the following occafion. 
Near to an iron rod, that touched the bottom of 
a jar charged politively, I placed another infulated 
rod, with a pair of pith balls hanging to it ; and 
obferved, that, when I attempted to make the dif- 
charge, through an imperfeClly conducting circuit, 
(bringing e , g> part of the table into it), a ftrong 
fpark palTed between the infulated rod and the other 
that touched the jar, and immediately the balls fepa- 
rated as far as they poffibly could ; and, continuing 
in a repullive Hate, appeared to be electrified nega- 
tively. But immediately completing the circuit with 
good conductors, and making the remainder of the 
explofion in a full fpark; another fpark palled 
between the two rods, and immediately the balls 
fell clofe together again ; and fometimes would fe- 
parate with the oppolite, i. e. pofitive, electricity. 
I could not, upon this occalion, make the lateral 
fpark, in the full explofion, fo great as in the im- 
i perfeCt 
