666 EXPERIMENTAL INVESTIGATION OF THE 
metals, it must have been in a state of peroxide,* 
and that more than an eight-hundredth part of the 
half-crown must have consisted of iron, if its mag- 
netism were due to the presence of that metal. 
62. Again: the magnetic action of this half- 
crown was considerably more feeble than that of 
‘the alloy which has been chemically examined ; 
and in which, if iron were present at all, that metal 
was in a less proportion than a twenty-thousandth 
part of the mass, which proportion, in a state of 
peroxide, and divided as it necessarily must have 
been through the whole alloy, would scarcely yield 
the slightest perceptible magnetic action. More- 
over, if iron to that amount were even pure or 
uncombined, its quantity was far too small for the 
display of those high magnetic powers of which it 
was obviously possessed. And, as there is a proba- 
bility, at least, that the magnetic powers of iron 
become deteriorated by an alloy of that metal with 
silver or with copper, or both, there is not the 
slightest reason for supposing that the magnetism 
of the alloy in question (32) was due to any iron 
that it could possibly contain. Nor do I believe 
* It is possible that the iron, if any, might be in the state 
of a carburet ; but even in this condition much of its magnetic 
powers would be neutralized. 
