[ 4-59 ] 
in pieces, the top falling from the bottom near the 
point of the lower wire. Another phial was fitted 
in the fame manner, and the cork cut longitudi- 
nally, that the air might freely pafs at the time of 
the explofion, but this made no fenfible difference : 
often times the phial is fo cracked as to refemble 
radii from a center. 
If oil is ufed inftead of water, the event will be 
the fame. 
The quantity of eleCtricity neceffary to burft the 
phial, appears to vary more in proportion to its 
thicknefs than its fize ; many phials of various fizes 
may be broken at io of the electrometer, while 
others, nearly of the fame fize, remain found, with 
a ftroke at 30, or even more. 
I generally found green glafs more difficult to 
break than white. 
When the phial is not broken by the eleCtric 
ftroke, the agitation of the water may be fenfibly 
obferved at the inftant of the explofion, and the 
eleCtric fpark evidently feen to pafs through the wa- 
ter, from the point of one wire to the other. 
This remarkable appearance of the eleCtric fluid’s 
pafling through water may be obferved, when the 
electrometer is at a fmaller diftance from the con- 
ductor, if the wires are nearer to each other. 
I have broken many phials by the eleCtric ftrokes 
as above-mentioned, when the wires have been at 
the various diflances, of above 1 inch to of an 
inch from each other, as near as my eye could de- 
termine ; but the diftance of about _L. of an inch I 
ufually prefer. 
N n n 2 The 
