Obfervations on EleSiricIty . 275 
it foon goes off, and is not fo ftrong as the foregoing, which 
lafts feveral days. 
23. To give fome diftimftive criterions by which other elec- 
tricians may determine whether the intenfity they produce ex- 
ceeds or falls fhort of that which this method affords, I fhali 
mention a few fafls. 
With a cylinder 7 inches diameter and cufhion 8 inches long, 
three brufhes at a time conftantly flew out of a three-inch ball in 
afucceffion too quick to be counted, and a ball of if inch dia- 
meter was rendered luminous, and produced a ftrong wind like a 
point. A nine-inch cylinder with an eight-inch cufhion occafioned 
frequent flafhes from the rourid end of a conductor 4 inches 
diameter: with a ball of 2f inches diameter the flafhes ceafed 
now and then, and it began to appear luminous : a ball of if 
inch diameter firft gave the ufual flafhes ; then, by quicker 
turning, it became luminous with a bright fpeck moving about 
on its furface, while a conftant ftream of air rufhed from it ; 
and, laftly, when the intenfity was greateft, brufhes, of a dif- 
ferent kind from the former, appeared. Thefe were lefs lu- 
minous, but better defined in the branches; many ftarted out 
at once with a hoarfe found. They were reddifh at the fiem, 
fooner divided, and were greenifh at the point next the ball, 
which was brafs. A ball of 4L inch diameter was furrounded 
by a fteady faint light, enveloping its exterior hemifphere, 
and fometimes a flafh ftruck out at top. When the excitation 
was ftrongeft a few flafhes ftruck out fideways. The horizontal 
diameter of the light was longeft, and might meafure one 
inch, the ftem of the ball being vertical* 
This laft phaenomenon is fimilar to a natural event related 
by M. Loammi Baldwin *, who railed an eledtrical kite in 
* Memoirs of the American Academy, Vol, I. p. 257. 
July,. 
