Oscillations of the Horizontal Needle. 21 
rapidly the arc, and to diminish the time of a given number of vibra- 
tions ; 2d, The influence of the sun’s rays on a magnet oscillating 
in a void, is to increase the time of a given number of vibrations, 
whilst the arc, if the retardation of the rate of vibration be small, 
is not materially affected. 
34. These phenomena, from the causes already considered 
(20) (24), seem quite consistent with each other. Thus, the 
more rapid diminution of the are in sunshine, may, in great mea- 
sure, arise from the state of the surrounding air, experiment (a) 
(4); since it is no longer apparent, in a strictly comparative case, 
when the air about the bar is made very rare; differences in the 
are might hence, in a perfect void, become indefinitely small. 
The differences which thus arise in the are, necessarily cause cor- 
responding differences in the rate of vibration, experiment (0) (4) ; 
and these will become more or less evident, according as they ex- 
ceed, or are less than, the effects of other causes, tending to a 
contrary result. 
The diminished rate of vibration under the influence of the 
sun’s rays in a void, may arise from expansion in the bar itself 
(20), or from a decrease in magnetic tension (24), or from both. 
If the magnetic state of the bar be at all constant, in regard to 
the small elevation of temperature to which it is exposed, the 
former would be the most probable cause of the effect, which, be- 
ing no longer masked by great differences in the are of vibration, 
becomes now very sensible. 
35. In the following Table, I have transcribed two series of 
experiments, carried on in vacuo, in which the time of 100 os- 
cillations is deduced from 150 oscillations observed in small ares. 
In these the retarding influence of the sun’s rays is also evident. 
