49 
Magnetical Experiments, fyc, 
a spare set of pins screwed throughout their length, and fut> 
nished with nuts. In addition to these bars, &c. I provided se- 
parate feeders or conductors of soft iron, suitable for connecting 
the poles of each of the bars of the compound magnety and also 
another conductor, fitted to the whole when combined. With 
this apparatus, I proceeded to give the magnetic virtue as fol* 
lows. 
I took a rod of soft steel, which I considered better than a 
poker, and hammered it for a minute or two, while held verti- 
cally upon a large bar of soft iron in the same position. This 
gave considerable magnetism to the steel-rod. On the top of 
this, I then hammered each of the six bars of soft steel, until 
the accession of lifting power ceased. Then fixing two of them 
on a board, with their different poles opposite, and formed* by 
a feeder at each end, into a parallelogram, I rubbed these, after 
the manner of Canton by means of the other four bars, and 
found their magnetism greatly augmented. The other four bars 
were operated upon in pairs, in a similar way, those already 
strengthened being used for strengthening the others, and each 
pair being successively changed, until all the bars were found 
to be magnetized to saturation. A pair of them now possessed 
a lifting power of two pounds and a half. 
The next step was to touch the bars intended for the com- 
pound magnet, by means of these six bars now magnetized. 
For this purpose, the six bars were combined into two magnets, 
by tying three of them together, with similar poles in contact ; 
these two were then placed, with opposite poles, in connection, 
and tied together at one end, but separated about the third of 
an inch at the other, so as to form one compound magnet, and 
a conductor was kept constantly applied to the open end of it, 
when not in use, to preserve the power from being lost. One 
of the bars of the horse-shoe magnet, with a conductor placed 
across the poles, was now placed on a board, in a groove cut 
out so as to hold it fast under the operation. The straight bar 
magnet was then placed erect on the middle of it, with the sepa^ 
rated poles downward, and rubbed against the horse-shoe bar, 
from the middle to one of its poles, until the north pole of the 
* See Phil. Trans., vol. xlvii. p. 31. 
VOL. IX. NO. 17 . JULY 1823 . 
» 
