556 GENERAL WALKER ON PENDULUxM OPERATIONS FOR DETERMINING 
pressure, 27"7 inches, the correction amounted to '53 under the temperature of 53° 
and to I'34 under the temperature of 100°. Corresponding corrections were therefore 
applied to the whole of the Indian swings, as they had already been provisionally 
reduced by the Kew formula. 
This must now be done for the revisionary swings at Kew and Greenwich, the 
results of which, as hitherto presented, have been reduced to the vacuum by the 
Kew formula only. It will be seen that the vibration-numbers at the pressure of 
27 inches are less than those at the pressure of 2 inches by 1'30 at Kew and 1”16 at 
Greenwich ; but it appears from Captain Basevi’s investigations that the high pressure 
results at both places should be increased by about O'7, which will reduce the 
discordances with the low pressure results to 0'60 and 0'4G. 
When pendulum operations are differential, and the variations of atmospheric pressure, 
at different stations, are small, the value of the correction for pressure does not require 
to be known with much accuracy. Beducing to a vacuum is also then unnecessary, 
and any convenient standard of atmospheric density may be employed instead. Thus 
Colonel IIerschel has sufficiently provided for the elimination of the effects of varia¬ 
tions of pressure at his stations, which were all of low elevation, by reducing his 
swings to the standard pressure of 26 inches under the temperature of 32° F,, instead 
of to a vacuum. 
The Indian swings were invariably reduced to a vacuum, in accordance with 
previous procedure. They were made in an exhausted receiver, under the lowest 
pressure attainable, partly because this enabled them to be extended over a longer 
time, and thus be less inffuenced by hourly variations of clock-rate, than if taken 
under full pressure ; partly because the receiver would protect the pendulum from the 
action of currents of air ; and partly to obtain as nearly a uniform pressure at all the 
stations as possible, and thus secure strictly differential results; for it was intended 
that the pendulums should be swung both at low levels in the neighbourhood of the 
ocean and at the highest attainable elevations, as on the tabledands of the Himalayan 
mountains, where the pressure of the atmosphere is halfway down to the vacuum, so 
that a considerable range of pressure had to be met with and provided for in the best 
way possible. By exhausting the receiver, the pressures under which the swings 
were actually taken were generally maintained between 1 and 2 inches, excepting at 
first, when slight leakage was met with, the locus of which was not immediately 
detected, and then the pressures ranged from 1 to 4 inches. But these differences of 
pressure are so small that the uncertainty as to the precise amount of the pressure 
correction cannot exert an appreciable influence on the differential results which have 
already been deduced, and which are the ultimate object of the observations. And 
this is the case also both in Colonel Herschel’s and in the revisiouary operations, the 
range of pressure being always under 2 inches, whether the swings were taken under 
high ]3ressures only or under both high and low pressures. 
Thus, a more exact knowledge of the correction for pressure might sensibly affect 
