6 Mr Harvey on the Influence of Magnetism 
rates may, however, be conveniently contrasted in a Table, with 
the results obtained during the first course of experiments with 
tlie same chronometei’c 
Situation of 
C imONOMETER (A). 
Rate. 
Rate. 
Situation of 
Chronometer (A). 
Detached, 
+ 20''.4 
+ 20" 1 
Detached. 
Fig. 1. (mean), 
+ C8''.9 
+ 50".8 
(Mean), Fig. 5. 
Fig. 2. 
— 23".2 
+ 29". 1 
Fig. e. 
Fig. 3. 
+ 43".4 
+ 33".7 
Fig. 7. 
Fig. 4. 
— 2".6 
+ 18".5 
Fig. 8. 
From a comparison of the two numerical columns of the pre- 
ceding Table, it is apparent that the position of the main-spring 
with respect to the magnet, has a very material influence on the 
rate of the chronometer ; and that the removal of the centre of 
the spring from the axis of the magnet, through equal arcs, ap- 
pears to have produced proportional declensions in the rate. 
Thus the rates corresponding to Figs. 1. and 3., being respec- 
tively -f- 68."9 and 43."4, are very nearly proportional to -|- 50."8 
and -{- 33."7, the rates belonging to Figs. 5. and. 7 No 
analogy, appears, however, to exist among the results obtained 
in the other positions of the chronometer. On the contrary, the 
difference in the declension of the rates obtained by changing the 
position of the time-keeper from Fig. 1. to. Fig. S., and from 
Fig. 5. to Fig. 6., is remarkable ; and particularly so, when it 
is considered, that the anomalous results are produced by an al- 
teration of only 27° in the position of the machine. The con- 
trast also in the rates obtained in the situations denoted by 
Figs. 3, 4, and Figs. 7, 8. is scarcely less singular. Kesults so 
exceedingly diversified could scarcely have been anticipated 
from so small an alteration in the position of the machine, and 
* This may be only an accidental coincidence. The close approximation, 
however, to proportionality in the numerical results, deserves some consideration. 
It would seem as if the translation of the chronometer from the position occupied 
in Fig. 1. through a given arc, occasions a declension in the rate proportional to 
that which takes place w'hen it is moved from the position denoted by Fig. 3.) 
( through an equal arc in a contrary direction. 
