462 
MR. BAILY ON THE CORRECTION OF 
observing and watching the motion of the pendulum during its vibrations: 
and I found that when the sphere was suspended by a wire over a cylinder, 
the motion of the ball, although set off in a straight line, soon became ellip¬ 
tical ; that the eccentricity of the ellipse was continually diminishing; and 
that the major axis was continually shifting its position with respect to the 
points of the compass: circumstances which were sufficient to account for all 
the appearances above described, and to destroy all confidence in experiments 
conducted in such a manner. And although I have retained the experiments 
with the pendulums No. 8 and 9, above alluded to, which were made in this 
way; yet it has been more to show the near accordance which may sometimes 
be accidentally attained by an incorrect method, and that we cannot examine 
too minutely into every step of so delicate an inquiry. 
I wish it however to be fully understood that these remarks do not apply to 
hi. Bessel’s experiments, since there is this important distinction to be made 
between his mode of proceeding and mine : viz. that his wire, at the part where 
it passed over the cylinder, was purposely made flat, probably with a view of 
avoiding this very difficulty ; whereas mine was round, as generally sold in the 
shops. I have not yet tried the flat wire, but have thought it right to point 
out the inaccuracies that may attend the use of the round wire, in order that 
others may not adopt it without the precaution of first ascertaining how far 
the results of any experiments may be affected by the anomalies above alluded 
to. In conclusion, I would add that, in the knife edge suspension, the vibra¬ 
tions of the ball were uniformly preserved in a straight line during the whole 
time it was in motion : and no anomalies were discoverable. 
Confined space of the Vacuum apparatus. 
It has been suggested by some persons that the results of experiments, of the 
kind mentioned in this paper, may probably be affected by the confined space 
of the tube in which the oscillations of the pendulum are made. M. Poisson, in 
his valuable memoir above alluded to, has justly stated that, in all the analy¬ 
tical investigations, the oscillations are supposed to be made in a fluid which 
extends indefinitely in all directions: a circumstance, however, which cannot 
practically take place in experiments of this kind. But he imagines that when 
the pendulum is small, in comparison with the dimensions of the inclosed 
