434 
MR. BAILY ON THE CORRECTION OF 
It appears from this Table that, in the case of spheres, whose diameters 
are rather less than H inch (which is about the size of that used by M. Borda, 
and by M. Biot, in their experiments on the length of the seconds pendulum), 
suspended by a fine wire, the value of n may in pendulums of such length be 
assumed equal to T86 : but that, if the diameter of the sphere be increased to 
about 2 inches, as in M. Bessel’s experiments, the value of n will be dimi¬ 
nished to 175 . I regret that my vacuum apparatus is so constructed that it 
will not admit of my making experiments on either larger or smaller spheres 
or on longer or shorter pendulums: otherwise I should have pursued this 
inquiry further, in order to discover the law by which the results of pendulums 
so constructed are governedIt will be seen likewise, from a comparison of 
the pendulums No. 10 and 13, that the size of the suspending wire, or rod, has 
a perceptible (although in those particular cases, not a very material) effect on 
the results : increasing the value of n, as the size of the wire increases. The 
value of n is affected also by the form of the rod, as may be seen by a compa¬ 
rison of No. 40 and 41, to which I shall again presently allude. 
The solid cylinder, 2 inches long, gives the value of n equal to J-86t; 
another, of the same diameter, and double the length, gives 2'03 ; and the 
cylindrical tube, 56 inches long, gives only about 2‘3 : whilst the small cylin¬ 
drical rod, not much more than 4 tenths of an inch in diameter, gives upwards 
of 2’9. Other apparent anomalies will present themselves, on a more minute 
examination and comparison of the values given in the Table; which can 
only be cleared up by future experiments. 
It appears also from this Table that the additional number of vibrations to 
be applied to the results from experiments with a platina sphere, similar to that 
made use of by M. Biot;};, will be 2-395 : whereas the additional number to be 
* I have made some alterations in my pendulum apparatus, since this paper was read, which has 
enabled me to extend the scale of my experiments; as I shall subsequently state more at length. 
f Since this paper was read before the Society, I have seen the account of M. Bessel’s additional 
experiments on the pendulum, in the Ast. Nach. No. 223. From those experiments, M. Bessel 
deduces the value of n, for a cylinder very similar to that mentioned in the text, equal to l - 755. In 
this experiment the length of the wire was nearly the same as mine. But, for his long pendulum, he 
makes the value of n equal to 1 '952. He has also slightly increased the value of n as adduced from 
his former experiments; making it equal to 1 '956, instead of 1 '946, as already mentioned in page 402. 
+ Although this is the value to be applied to the pendulum used by M. Biot, it does not follow 
that it would be correct to apply the same value to that used by M. Bokda (which was a two seconds 
pendulum), unless it should be found that the factor is the same for long and short pendulums of this 
construction. 
