1822.] Philosophical Transactions for 1822, Part II. 457 
filing, bending, &c. are governed,” and which have been pub- 
lished in the Edin. Phil. Trans. for-1821, confines himself, in the 
present communication, “to the application of these laws to 
practical magnetism; and particularly to the construction of 
magnets, without the use of any magnetized substance.” 
_ “ In examining the magnetical effect of percussion on different 
kinds of iron and steel, two tests were employed ; the weight of 
iron that the body would lift, and the quantity of deviation that 
it would produce on a magnetic needle when presented to it in a 
certain position, and at a certain distance. For the first test, 
common iron nails of different sizes were made use of : they were 
of the weights of 2, 4, 64, 14, 24, 37, 45, 88, 130, and 188 ers. 
For the purpose of securing a good and uniform contact with 
the magnetized bar, the oxide on the ends of the nails was 
removed by means of a fine file, and ihe extremities were then 
polished by rubbing them on a Turkey stone. The second test 
i employed consisted of a board two feet in length, with alongi- 
tudinal line down the middle divided into inches, anda sensible 
pocket compass. To guard against the effects of the magnetism 
of position, the central line of the board was placed exactly in an 
east and west direction by the compass; and as the board was 
laid horizontally on a table, this line was known to be in the 
plane of the magnetic equator, and consequently in a situation 
in which small bars of iron are not affected by the magnetism of 
position. In applying this simple apparatus as a measure of 
magnetism ; the bar, whose magnetism was to be examined or 
compared, was laid along the central line of the board, with its 
north pole always nearest the compass. The compass was 
placed with its centre at the commencement of the scale, so 
that its needle was exactly at right angles to the direction of the 
bar ; and before the deviation took place, its poles were equi- 
distant from the bar. The distance was estimated by the scale 
on the board, and always represented the space between the 
north end, or nearest end of the bar, and the centre of the com- 
pass. Three hammers were also employed: No. 1, of 22 ounces; 
No. 2, of 12 ounces; and No. 3, of 24 ounces weight. 
With this apparatus, a number of experiments were performed, 
several tables of which are given: and their general results are 
stated as follows : 
“]. A cylindrical bar of soft steel, 64 inches long, and 
weighing 592 grains, lifted, after repeated hammering on pewter 
and stone, 6} grains; but could not be made to lift a nail of 11 
rains. - 
«2. The same bar hammered vertically upon a parlour poker, 
also held erect, after 22 blows, lifted with the lower end, 
which was a north pole, 88 grains ; and on using a larger ham- 
mer, received a considerable increase of power, producing a 
deviation of the compass, three inches distant, of 34 degrees: 
further hammering, it was found, rather diminished than increas- 
