MAGNETIC CHARACTERS OF METALS, &c. 665 
the latter are affected by its union with those 
metals, or not, is not yet known. 
60. In order to ascertain the exact quantity of 
pure metallic iron, that would render a neutral 
half crown* apparently magnetic, to the same 
extent as another half-crown, was absolutely 
magnetic, I attached to the former, by means of 
softened gum, new iron filings; and after many 
trials ascertained that the requisite quantity of 
filings amounted to about a ten thousandth part 
of the mass. And on changing the pure iron for 
the peroxide of iron, a 480th part of the mass 
was required to render it equally magnetic with 
the standard half-crown. Now, as more than 
two-thirds of this oxide is iron, it follows that, in 
this state, the iron loses a considerable portion 
-of its magnetic powers; and that the proportion 
of iron required in this case, to produce the 
standard degree of magnetism to the mass, was 
little short of a seven hundredth part. 
61. Now, if any iron existed in the standard 
magnetic half-crown, uncombined with the other 
* This piece of coin was not entirely devoid of magnetic 
action, but it approached nearer to a state of neutrality than 
any other I then had. Its action was very feeble indeed, 
