64 Mr. Cavallo’s Magnetkd 
feme cobalt was contained in them ; but I fee no reafon why 
the nickel, when alloyed with a little cobalt, fhould fhew no 
magnetifm, if that property does really belong effentially 
to it. 
The greateft number of my experiments are relative to the 
properties of brafs ; and they feem to prove, that this com* 
pound metal, which is often magnetic, does not owe its mag- 
netifm to iron, but to fome particular configuration of its com- 
ponent particles, occafioned by the ufual method of hardening 
it, which is by hammering- 
In fome fpecimens of brafs, and efpecially in that which has 
often palled from the work-lhop to the furnace, and from the 
latter to the former, there are fometimes pieces of iron fenfible 
not only to the magnet, or the chemical analyfis, but even to 
the fight, which render the brafs ftrongly magnetic. But the 
brafs generally ufed in my experiments was fuch as, when quite 
foft, it had no fenfible degree of magnetifm. 
Before we begin with the narration of thofe experiments, it 
will be proper to deferibe the magnetic needle I generally ufed, 
which is fufpended in a particular manner ; and which may be 
ufeful to perfons who are fond of making magnetic experi- 
ments, not only for its fenfibility, but likewife for the fimpli- 
city of its conftru&ion. 
Experience having fhewn, that large magnetic needles are 
not proper for experiments wherein a very frnall degree of mag- 
netifm mu ft be afeertained, and the free motion of the ufual 
frnall needles being proportionally more obftrufted by the nature 
of their fufpenfion, even when furnilhed with agate caps, I 
endeavoured to contrive a fort of fufpenfion which might an- 
fwer the purpofe better than the needles fufpended in the ufual 
manner; and, after feveral attempts, at laft I conftrufted a chain 
of 
