544 GENERAL WALKER ON PENDULUM OPERATIONS FOR DETERMINING 
Colonel Herschel got over this difficulty, and eliminated the influence of hourly 
variations of clock rate, by linking successive sets of swings together so as to flil up 
the whole 24 hours. He made the duration of each set of his swings somewhat less 
than 6 hours, taking all necessary observations of temperature, coincidence, &c. at the 
commencement and conclusion of the set, but without any intermediate observations; 
imniediately after the conclusion of one set he commenced the next set. In this way 
observations were sometimes carried on continuously for two or three days by himself 
and his assistant. In the revisionary operations at Kevv a similar procedure was 
adopted ; the temperatures and coincidences in the intervals between the beginning 
and end of each set were generally observed also, and the temperatures when not 
observed were recorded on a thermograph. At Greenwich observations were made at 
10 A.M., 4 P.M., and 10 p.m. ; the swings at low pressure were divided into two sets of 
6 hours each and one set of 12 hours, so as to fill up the 24 hours; those at high 
pressure into two sets of 6 hours each, to fill up 12 hours. The temperatures between 
the thermometer readings were registered by a thermograph. The daily range of tem¬ 
perature was very small at Kew, and rarely as much as 1° F. ar Greenwich. At both 
places the time was derived electrically from the siderial standard clock at Greenwich, 
which is fixed in the basement of the Observatory, where there is no sensible diurnal 
variation of temperature. Under the actual circumstances there was no real necessity 
for continuous observations throughout the 24 hours to control the clock rate. 
The differences in procedure and manipulation which have been pointed out thus 
far are not such as to have affected the results sensibly; but in one other matter 
there was a difference of procedure vdiich might have materially influenced the 
results. Colonel Herschel did not maintain the agate planes in exact horizontality; 
he believed that when the planes were truly level, and the pendulum was set up on 
them, the knife-edge, if pressed down against them hy hand, was found to be not truly 
in contact with the planes throughout the line of bearing ; consequently he dislevelled 
the planes until the contact, as judged by touch, was thorough, and then he commenced 
swinging the pendulum. Such imperfect contact was never noticed in the Indian 
operations; it is possible with pendulums having a rigid bar, when the knife-edge is 
not truly perpendicular to the bar; but with pendulums such as these, which have a 
very flexible bar, it seems scarcely possible, at least without a grosser displacement of 
the knife-edge from the perpendicular than is at all probable. The officers of the 
United States who swung these pendulums at several stations have been questioned 
on this point, and Mr. Edwin Smith reports that after levelling he “ tested the 
contact of the knife-edges with the agate planes by touch and found it impracticable 
to make any change, so the pendulums were always swung with the planes levelled 
with the spirit level.” In the revisionary swings this was done also. 
Colonel Herschel does not appear to have measured the actual dislevelment of the 
agate planes which was caused by his method of treatment; had he done so and his 
surmise been correct, the magnitude of the dislevelment would have been the same 
