to ascertain a Standard of Weight and Measure . 161 
should be such as to amount to y^oo i nc ^> anc * these so many, 
without balancing each other, as to produce a spheroidical form, 
one of whose diameters should exceed the other by — J oo inc ^ 5 
in that case, the error in the assumed solid would not exceed 
__ J __ part of the whole ; and this is a position infinitely too ex- 
travagant to be admitted, when we recollect, that this diameter 
has been probably taken to within a tenth part of that error. 
(§. 29.) The weight of this sphere, in air and in water, comes 
next under our consideration ; the experiments for which were 
as follow, made June 12, 1797; the barometer being at 29,74 
inches, and the thermometer, in air, at 67°. 
Experiment the 1st. 
The weight of the sphere in air, the counterpoise,"] 
or weight of the scale or cradle, a bcf, (Tab. VII. I 
fig. 3.) in which the sphere hung, being allowed j ' 
for*, so that this was the net weight - J 
The sphere and scale suspended in 1 grains, 
water, with its centre 5,6 inches 1= 303,17 
below the surface, and the heat 66° „ 
Deduct the counterpoise, or weight! 
of the scale, in water, with the 1 ___ 
same heat of 66°, and same depth-f' | 2 53 >o 2 
below the surface - J 
The difference is the net weight of the sphere in 
water, of the temperature 66°, which, deducted 
from its weight in air 
Leaves the weight of a bulk of water = the sphere, 
in the temperature 66°, and 5,6 inches below 
the surface - 
Troy grains. 
28722,64 
— 49. 8 5 
28672,79 
* The weight of this scale, with its 3 wires, in air, was rr 276,10 grains, 
f The sphere having been weighed in the same depth of water that the counter- 
poise to the scale was determined in, no correction for the greater or less immersion of 
the scale-wires was here necessary ; which however will sometimes be the case. See 
§. 29. and table II. of correction, §. 23. 
MDCCXCVIII. Y 
