[ 335 ] 
cushion without performing a revolution with the 
glafs j and that, of courfe, a circulation of the tire 
is thus kept up in the fubftance of the cufhion in the 
common method of conftruCting the machines. 
To be certainly convinced of this, I attempted to. 
make the p adage of the fire from the glafs to the 
anterior part of the cufhion, or to that part whfch. 
correfponds with the afcending fide of the cylinder, 
demonftrable, by placing a piece of filk between, 
the glafs and cufhion. This filk was larger than 
the cufhion ; and part of it was allowed to adhere,, 
by the attraction of the eleCtric fire, to the afcend- 
ing part of the cylinder.. My view in doing this 
was to cut off, in that part, the immediate com- 
munication between the excited glafs and cufhion, 
and by that means render the circulation of eleCtric 
matter vifible,. which I fufpeCted to take place in 
the machine ; as it was thus forced to turn over the 
loofe edge of the filk before it could return to the 
cufhion. The event anfiwered my expectation ; 
and 1 then perceived, that the greateft part of the 
excited fluid was commonly re-abforbed by the fore- 
part of the cufhion without becoming fenfible cn, 
the fuperior part of the glafs.. 
Having thus verified my fuppofition by aCtual; 
experiments with filken flaps of different fizes, I 
endeavoured to difeover a method of preventing that 
circulation of the eleCtric fluid, and,, if poffible, 
cf obliging the whole, or the greater part of it, that 
is once excited, to make the revolution with the, 
glafs. This, indeed, the filk, when of confide* 
rable breadth, in fome meafure effected ; but I 
thought that this obftruCtion to the immediate re~ 
5 turns 
