Mr. Bennet’s Defcnptlon , &c. 295 
The atmofpherical electricity was fometimes fo ftrong as to 
need no doubling, and moftly required only one application of 
the fecond plate, yet I frequently found it neceflary to repeat 
the procefs from two or three to twenty times. Perhaps the 
exaCt comparative quantity of eleCtricity redding in the atmo- 
fphere might be meafured by the number of operations re- 
quired to render it perceptible by the electrometer, all other 
circumftances being cautioufly attended to. 
If the eleCtricity of the atmofphere (hould happen to be 
much weaker than I have yet found it, there remains not only 
the refource of doubling oftener, but the capacity of the in- 
ftruments may be much increafed; as, firft, by ufing a larger 
flame ; fecondly, by elevating it higher ; thirdly, by collecting 
the eleCtricity with a very thin glafs ball, filvered within, and 
coated on the outfide in the common way, or gilt ; fourthly, 
by grinding and polilhing the plates of the doubler very 
exaCtly ; fifthly, by making the experiments in an advantageous 
fituation. In all thefe particulars my apparatus was defective, 
yet amply iufficient for the difeovery of the atmofpherical 
eleCtricity. 
After confidering the fuccefsful effeCt of flame, in collecting 
atmofpherical eleCtricity, I placed an infulated lantern upon a 
pole about fifteen feet high, and fufpending a gold thread from 
the lantern connected it with the electrometer, and was agree- 
ably amufed with feeing the pendulous gold leaf open and (hut 
with every pafling cloud. 
On the 27th of February, 1787, when there was a confi- 
derable mift whilft the lantern was thus elevated, the leaf 
gold frequently (truck the fides of the electrometer ; and, in 
Vo l . LXXVII. X x about 
