664 EXPERIMENTAL INVESTIGATION OF THE 
generally, as it leaves the foundry, contains any 
notable quantity of iron. I have analysed many 
specimens, both magnetic and unmagnetic, and 
the traces of iron, where any were discoverable, 
were very minute, and as frequent in the one kind 
as in the other. It is true that some specimens 
of brass contain more than an average proportion 
of iron; but it is a curious fact that these speci- 
mens are not those which display the greatest 
magnetic powers. 
57. We learn, also, from these facts, that the 
demagnetizing powers of antimony and zine will 
necessarily prevent the detection of small pro- 
portions of iron in those metals, even by the aid 
of powerful magnetic forces, and leave us in 
uncertainty regarding their purity when examined 
by this test alone. 
58. Lead and iron do not easily unite into a 
perfect alloy, excepting when the ferruginous 
metal is in very small proportions; but when 
thus combined, the iron loses a great part of its 
natural magnetic qualities. 
59. Silver and copper unite very sparingly 
with iron; but whether the magnetic powers of 
