562 
SECOND REPORT— 1832 . 
is inclosed in appropriate receivers. The bar is suspended in 
the usual way, by one or two slight filaments of silk, about 15 
inches in length, which is attached to a vertical and sliding rod 
of brass, so as to admit of being raised or depressed, through 
a small distance, to any required point. 
When the exhaustion is complete, the air-pump barrels may 
be taken away, if required. 
The whole of the above-mentioned apparatus may be taken 
in pieces, for the convenience of carriage, without difficulty. 
The possibility of change in the magnetic state of a bar, has 
necessarily occupied the attention of those who have at various 
times been engaged in this department of science,—the iden¬ 
tity of the magnetic pendulum not being preserved, if its mag¬ 
netism should vary either in intensity or distribution: it has 
hence been proposed, as a means of discovering whether any 
change has taken place in the magnetism of a bar, to return 
again to the first place of experiment, and observe whether the 
same rate of vibration takes place, as had been observed in a 
previous instance, under the same circumstances. In the pre¬ 
sent state of magnetic inquiries, however, this method is by no 
means free from fallacy, since we do not know of what change 
a magnetic bar may be susceptible, in consequence of a change 
of place on the earth’s surface; nor are we quite certain, that 
the directive force of the earth is, in the same place, an invari¬ 
able quantity for all periods : it seems, therefore, desirable to 
obtain some method of detecting changes in the magnetism of 
a bar, which may be available at any time and in any place ; as 
also a method by which its magnetic state may be rendered 
invariable, at least as far as the operation of ordinary causes of 
disturbance interfere. 
The author proposes, as a delicate test of change in the 
magnetism of a bar, the oscillations of a ring of copper about 
its poles, the bar being fixed in the direction of a diame¬ 
ter of the ring, with its ends at about the y^th of an inch 
distance from the inner surface of the ring. For this purpose, 
a light ring of copper, about half an inch in depth, and Jth of 
an inch in thickness, is suspended over a graduated cord by 
two parallel threads, without torsion; the threads being fixed 
to a cross bar, which is a diameter of the ring, and at about 
^th of an inch on each side of the centre. The ring is fur¬ 
nished with index points, and is to be suspended in the frame¬ 
work of wood of the apparatus above mentioned ; the cross-bar 
is turned to any required angle, from the direction of the paral¬ 
lelism of the threads, in which the ring remains at rest by the 
same means as are employed to deflect the needle, and again set 
