8 Mr Harvey on the Influence of Magnetism 
+ 8". 5, agreeing within a second of its former rate in the same 
situation. The comparative influences of the two poles may 
be readily observed in the following Table : 
Situation of Chronometer (B). 
Applied to the 
North Pole of 
the Magnet. 
Applied to the 
South Pole of 
the Magnet. 
Fig. 10. Main-spring nearly in contact with 
the magnet, and the attractive power trans- 
mitted through its centre, 
+ 11^7 
+ 9".5 
Fig. 11. Centre of the main-spring 90® from 
the preceding position, and the magnetic 
power transmitted nearly through the cen- 
tre of the balance and its spring, the balance 
being at its least distance from the Pole, 
-j- 3".4 
— 4".2 
Fig. 12. Centre of the main-spring 180* from 
its first situation, and the magnetic power 
transmitted through its centre, 
4- 7''.0 
4 8".7 
Fig. 9. Centre of the main-spring 90® from its 
first situation, and the magnetic transmit- 
ted nearly through the centre of the balance 
and its spring, the balance being at its 
greatest distance from the Pole, 
+ 1".4 
— 0".8 
It may be proper, however, to remark, that at the time the 
chronometer (B) was successively applied to the south pole of 
the magnet, another chronometer (C) was placed at its north 
pole, as represented in Figures 9, 10, 11, and IS. ; and that a 
series of experiments, similar to those already detailed, pro- 
duced an effect on the last mentioned machine precisely the re- 
verse of that which had been observed in chronometers (A) and 
(B). Thus, when the latter chronometer was applied to the 
south pole of the magnet, as in Fig. 9., the mean of two sets of 
observations gave a result of — 0''.9 ; but a mean of two sets 
of chronometer (C), when applied to the other pole, furnished 
an average rate of -f 15''''.7 ; the detached rate of the former 
machine having been — and of the latter -f S".9 ; the 
positions of the balance and main-spring being precisely similar 
in both cases. In like manner, when the chronometers were 
posited, as in Fig. 10., the chronometer denoted by (B) gave a 
result of -f- 9"*5, at the same time that (C) furnished a mean 
rate of — 4".2 ; and when the former produced an average rate 
of — 4".S, in the situation represented in Fig. 11. (the perfect 
coincidence of the rates of two chronometers so differently cir- 
