xvith an Invariahk Pemlulmn. 
The length of the seconds pendulum at Rio de Janeiro isj S9, 04381 
Diminution of gravity from the Pole to the Equator, . .0053365 
1 
Mean Ellipticity^ ^ ^ . * r ° 301 77 ”^ 
Length of equatorial pendulum, » = 39,01223 
From another set of observations, ma'de at the same place, 
and with the same pendulum, by Mr Henry Foster, the follow- 
ing results were obtained : 
N umber of vibrations, ^ 
Length of pendulum, 
Diminution of gravity, * 
Mean ellipticity, . , 
Length of equatorial pendulitm, 
86131,55 
39,04368 
.0053431 
1 
302,37 
39,01206 
Captain Hall considers these results as more entitled to credit 
than his own, as the temperature was more steady, and the 
clock’s rate more uniform. 
The attentive reader will already have remarked, that the in- 
variable pendulum with which these observations were made, is- 
said to have performed 86235,98 vibrations at London, in 
1820 ; v/hereas, when it was brought back, in 1823, it performed 
86236,95 vibrations, being a dilference of 0.97. As this diffe- 
rence could not possibly arise from errors of observation, it be- 
comes highly interesting to discover the cause of it. 
Captain Kater,” says Captain Hall, was disposed to assign 
it to an accident v/hich had happened to the pendulum at San 
Bias, but which I, at first, imagined inadequate to such an effect. 
The accident was this : The pendulum, when not in use, was, 
as' usual, raised by means of a screw, so that the knife-edge was 
lifted clear of the agate planes on wdiich it vibrated during the 
experiments. This screw being too small, or having some flaw 
in it, unexpectedly broke at San Bias, before the experiments 
there were begun ; and although the knife-edge was not raised 
more than the twentieth of an inch, yet, as the pendulum weighed 
more than 15 lb., the fall might, he thought, have altered the 
form of so delicate an edge in a slight degree, and thus have 
virtually lessened the distance between the point of suspension 
and the centre of oscillation 
* If the knife-edge be supposed to have become cylindrical, the virtual point cf 
suspension, it has been demonstrated, would be at the distance of the radius of 
curvature of this cylindrical portion below its surface, and the number of vibrations- 
of course be greater than before.” 
