216 
CATTAIN SABINE ON THE REDUCTION TO A VACUUM 
at 72°, is 5.S8 vibrations. The difference, or 3.845 vibrations per diem, is the 
amount by which the experimental reduction to a vacuum exceeds the reduction 
which it has been customary to compute. 
From the imperfect state of the apparatus in this first experiment, doubts 
might have been entertained of the correctness of the result on two accounts : 
it might have been supposed, 1st, that the abstraction of the air being kept up 
continually, during the vibration in the rarefied medium, to counteract the 
leakage, currents might have been occasioned influencing the time of vibration : 
or, 2nd, the iron bars supporting the pendulum not having sufficient spread 
at the bottom to counteract the lateral force arising from the vibration, and 
the point of suspension itself partaking of it in consequence, it might have 
been supposed that the time of vibration was unequally affected thereby in the 
air and in the rarefied medium. By experiments made with the same pendu- 
lum on the 8th and 9th of July, an account of which is already before the 
Society, (Phil. Trans. 1829, Art. IX.) in which experiments the pendulum was 
suspended from Captain Rater’s original mahogany support in the same room, 
the vibrations on an immoveable support, all other circumstances being the 
same, were found to exceed those on the plate of the vacuum apparatus, by about 
18 vibrations a day; due, doubtless, to the motion of the plate during the 
vibration, arising from the elasticity of the iron bars and their insufficient 
spread. To give more firmness to the suspension in the vacuum apparatus in 
a second experiment, inch boards of well seasoned oak were inserted verti- 
cally, having their lower ends resting on the interior of the iron cylinder 
which supports the glasses, and the plate was screwed down firmly on their 
upper ends by screws working into the iron bars : the suspension plate was 
thus directly and firmly connected with the foot cylinder by means of the 
boards independently of the bars ; the boards being hollowed out in the proper 
places to admit the observation of coincidences. 
To detect where the leakage took place, the interior of the apparatus was 
filled with water as high as the lower glass cylinder, and a communication 
being established between the exhausting pipe and the upper part of the in- 
terior, the air was withdrawn ; when bubbles of air were seen to rise rapidly 
from the interior surface of the iron foot cylinder, particularly from those parts 
of it which were opposite the flanches on the outside, where the metal was 
