20 Mr. Cavallo’s Obfervations '■n 
preceding obfervation was one hour and a quarter ; after which 
the electrometer remained ienfibly electrified for a long time. 
In repeating this experiment, the times elapfed between the 
correfponding oblervations did not follow ftrictly the fame pro- 
portion of increafe ; nor did they increafe regularly in the fame 
experiments, which may be attributed in great meafure to the 
inaccuracy in obferving, and to the fluctuating ftate of the air ; 
but it could be fafely inferred from all the experiments, that 
the times required for the dilperfion of the eleCtricity were at 
leaft greater than the inverfe duplicate proportion of the den- 
sities of the eleCtricity remaining in the electrometer. * And if 
we imagine, that they continue to diminith in the fame propor- 
tion of increafing time, which is far from being an extrava- 
gant fuppofition, we fhall find, by a very eafy calculation, 
that about two years after the electrometer would ftill retain the 
hundredth part of the eleCtricity communicated to it in the 
beginning of the experiment; and as we do not know how 
far a quantity of eleCtricity is divifible, or to what extent it 
may be expanded, we may conclude with faying that, ftriCtly 
fpeaking, the electrometer would remain eleCtrified for many 
years. 
It may be inferred from this, as well as from many other 
experiments, that the air, or in general any fubftance, is a 
more or lefs perfeCt conductor of eleCtricity, according as the 
eleCtricity which is to pafs through it is more or lefs condenfed ; 
fo that if a given quantity of eleCtric fluid be communicated 
to a fimall brafs ball, one may take it away by Amply touching 
the ball with a finger ; but if the fame quantity of eleCtric 
fluid be communicated to a furface of about ioo or 1000 
fquare feet, the touching with the finger will hardly take away 
any part of it. 
7 
If 
