566 
SECOND REPORT -183/2. 
being wholly unchanged ; at least the difference is so little, as 
not to be worthy of notice. 
The rays of bright sunshine, having been thrown at first 
upon the bar in a direction from south to north, they were sub¬ 
sequently thrown upon the bar in the reverse direction, from 
north to south, but without any difference in the result: hence, 
it does not apear that the direction in which the rays fall on 
the needle, in respect of the magnetic meridian, is of any con¬ 
sequence. 
Whatever may be the value of the above results, when taken 
as evidence against the opinion that the sun’s rays possess any 
magnetic influence, yet the author considers that they are fully 
competent to elucidate the phsenomena which ensue in ob¬ 
serving the oscillations of a magnetic bar, in the shade and in 
the sunshine. 
Thus, the more rapid diminution of the arc of vibration in the 
sun’s rays in air, is dependent on the change effected in the 
elasticity of the air surrounding the bar: since the arc is not 
similarly affected, in a strictly comparative case, when the air is 
greatly rarefied. 
The increase of the rate of vibration in air, in the sunshine, 
necessarily ensues in consequence of the great differences in the 
arcs of vibration, which, in the sunshine, become more rapidly 
diminished ; this effect will be more or less prominent, in pro¬ 
portion to its exceeding or falling short of the effects produced 
by other causes, tending to an opposite result. 
The slight decrease in the rate of vibration, in a rare medium, 
is referable to the effect of expansion on the bar itself, by which 
its length, as a pendulum, becomes changed; and which effect, 
being no longer masked by the differences in the arc of vibra¬ 
tion, as in the former case, becomes now apparent. 
For the purpose of inquiring more minutely into the accuracy 
of the foregoing deductions, the following experiments were in¬ 
stituted. 
1. A copper bar, of the same dimensions as the magnetic 
bar, was set vibrating by means of a suspension with two pa¬ 
rallel threads, as explained in the early part of the paper ; and 
the times and arcs of vibration noted as before, both in air and 
in an exhausted receiver : the results were precisely the same, 
except that the slight retardation in the rarefied medium was 
not very appreciable. In this case, there was no magnetism 
proper to the bar; hence these effects are not exclusively de¬ 
pendent on that, agency. 
2. The oscillations of the magnetic bar, employed in the 
previous experiments, were observed for 25 minutes, during a 
