[ 442 ] 
I would alfo recommend rhe life of fuch to others, 
who (hall pleafe to employ themfeives in the like 
enquiries. Thofe I have provided for myfelf have 
a Troy Ounce for their integer , and my lead: weight 
is the thoufandth part of that quantity, differing 
confequently from the half of a Troy Grain only 
as 24 does from 2y, which is inconfiderable fo far 
as thofe fmall weights are concerned. My four 
fmalled are refpedtively of 1, 2, 3 and 4 of thole 
thoufandth parts, and together make 10, or an 
unit of the next denomination, that of the ico th 
part of an ounce. I then have four others, making 
1, 2, 3 and 4 ioofo*, and together the unit of 
of the next denomination, or one tenth of an 
ounce, and fo on. By thefe I five the trouble of 
reducing the common weights to their lowed de- 
nomination in every experiment, and fometimes 
perhaps avoid making midakes in that very trifling 
work. 
Whenever two or more original writers nearly 
concur in their experiments upon any fubjed, the 
Gravity fo deduced may be well depended upon. 
But where they differ remarkably it mud either be 
imputed to the unequal gravity of the fubjed itfelf, 
or to fome error in the tryals, which may eadly 
happen in matters that depend on the obfervation 
of fo many minute particulars. All thofe cafes that 
fo fenfibly differ would well deferve to be re- 
examined. 
The did Table above, that of Metals, as it is 
compofed of the mod perfed and uniform bodies 
in nature, feems capable of being adj uded with the 
great-eft precifion, both with relation to the pure 
Metals 
