[ +54 ] 
an inch, a very fmall ftroke will be perceived, with 
an explofion from K to M; and if K and M are 
diftant one inch from each other, the quantity of the 
dearie fluid, at the time of the explofion, will be 
increafed too times : for example, it appears by 
experiment, that, if the explofion happens after 4 
turns of tjie wheel B, when M is diftant from K 
t*t of an inch, or 1 turn of the ferew j the fame 
will happen at 8 turns of the wheel, when M and 
K are diftant 2 turns of the ferew, or of an 
inch j and if K and M are diftant 3 turns of the 
lerew, the turns of the wheel will be 1 2 at the time 
of the explofion ; the fame proportion will conti- 
nue fo far as the diftance of K and M is equal to the 
condenfing power of the coated phial without 
wafting. By wafting, I mean when the phial is fo 
fully charged, that part of the eledtric fluid efcapcs 
from the mouth of the bottle, or from the conductor 
into the air, or to fotne adjacent non-eledtric. The 
number of turns of the wheel, when K and M are 
at any of the above diftances, will be more or lefs in. 
proportion to the ftate of the air, the cylindrical glafs, 
the cufhion againft which the glafs is rubbed, or 
the coated phial ; which laft will not give fo great an 
explofion when the air is damp as when dry. 
The fewer the number of turns of the wheel, at 
any given diftance, the better the machine worketh. 
Thus a comparative difference between any two 
machines may be determined. 
A wire in general is better than a chain, unlefs 
the chain is held very tight ; particularly in very 
fmall ftrokes, the eledtric fluid will be loft in palling 
from link to link of the chain. 
By 
