in the Length of the Pendulum Vibrating Seconds, 3S3 
From the experiments given in the former part of this Re- 
port, it appears probable, that if the uncertainty which must 
exist in the allowance for the height above the level of the sea 
be excepted, the error in the number of vibrations of the pen- 
dulum, at any particular station, does not amount to so much as 
one- tenth of a vibration, which is nearly equivalent to 
part of the length of the seconds pendulum. To this degree of 
accuracy consequently may gravitation be determined by the 
apparatus I have employed ; and in passing through a country 
composed of materials of various densities, the pendulum may 
be expected to indicate such variation with very considerable 
precision^ 
The diminution of gravity from the Pole to the Equator is 
derived from the decrease which is observed to take place be- 
tween any two given latitudes ; consequently, if no irregular at- 
traction occurred, the results, computed from different portions of 
the meridian, should be the same. But it may be seen in the 
preceding table, that the number expressing the diminution of 
gravity, from the observations at Unst and Portsoy, is less than 
that deduced from the arc between Unst and Leith, and that 
this number goes on increasing to Clifton, diminishes at Arbury 
Hill, and increases again at London. It may also be remarked, 
that the diminution of gravity, derived from Unst and Dun- 
nose, is less than that deduced from Portsoy and Dunnose; 
from all which it seems probable, that in advancing south- 
ward, gravity decreases more than it ought to do from theory ; 
that there exists an assemblage of materials of greater density 
than common in the vicinity of Portsoy, and that the density of 
the strata to the southward becomes less and less until we arrive 
at Clifton, where it seems to be considerably in defect. 
At Arbury Hill, a sudden increase of gravitation is percepti- 
ble, and at the ^hort distance of London, this additional force is 
no longer sensible. From its intensity, and the limited sphere of 
its action, it might perhaps be inferred that the disturbing ma^ 
terial is of considerable density, and not very distant from the 
surface. 
It must be evident, that nothing very decisive respecting the 
general ellipticity of the Meridian can be deduced from the pre- 
sent experiments. For this purpose, it is requisite that the ex- 
