A PENDULUM FOR THE REDUCTION TO A VACUUM. 
425 
vity of the moving mass was reduced to 2‘079: and this is the value which 
M. Bessel employs, in deducing the results from the first set of experiments, 
where he swings the cylinder (closed) first in air and afterwards in water: 
which result gives n — 1*754. In the second set of experiments, he takes 
away the bottom piece of the cylinder, and having swung it first in air (where 
the diminished specific gravity was nearly the same as before), then immerses 
it in water, whereby, he says “ the specific gravity of the brass, about 8*3, 
“ is restored.” With this assumed specific gravity, the value given by M. 
Bessel is certainly correct. But if we suppose that the specific gravity of the 
moving mass is not restored to the specific gravity of the metal by suffering 
the tube to be filled with water; and that the pendulum can be considered in 
no other light than as consisting of a cylinder filled with water, instead of a 
cylinder filled with air ; (the specific gravity of which, instead of being 8*3, will 
probably not be so much as 2*8;) the value of the result will be materially 
altered. In fact, if the specific gravity were only 2*5, the value of n would be 
only 1*85 : which differs very little from the value deduced by M. Bessel from 
the experiments with the closed cylinder. Now, I find (from the data furnished 
by M. Bessel,) that the specific gravity of the cylinder when filled with water 
and with the bottom piece annexed, is about 2*8 : but it is evident that, when 
the bottom piece is taken away, the specific gravity will not be so much ; and 
by the assumption of such diminished specific gravity, the discordancy noticed 
by M. Bessel, would be considerably reduced, if not wholly eliminated *. 
Fifth set.— Results with the 2-inch Leaden Lens. 
No. 20. 
Exp. 
n 
105—106 
107—108 
Mean = 
1-614 
1-546 
1-580 
* M. Bessel remarks that in this experiment there was a more than usual motion of the water, 
arising from a portion of the fluid flowing out of the cylinder to supply the vacuum caused by the 
motion of the cylinder; and the reverse. But the effect of this on the general result would, I appre¬ 
hend, be very slight. In my experiments with hollow cylinders, above detailed, we observe but a 
trifling difference when the ends of the cylinder are left open. 
