'Experiments and Obfervatlons . yr 
jdly, Thofe pieces of brafs which have that property, retain 
it without any diminution after a great number of repeated 
trials, viz. after having been repeatedly hardened and foftened. 
But I have not found any means to give that property to fuch 
brafs as had it not naturally. 
4thly, A large piece of brafs has generally a magnetic power 
fomewhat ftronger than a fmaller piece ; and the flat furface of 
the piece draws the neeedle more forcibly than the edge or 
corner of it. 
5thly, If only one end of a large piece of brafs be ham- 
mered, then that end alone will difturb the magnetic needle, 
and not the reft. 
6thly, The magnetic power which brafs acquires by ham- 
mering has a certain limit,, beyond which it cannot beincreafed 
by farther hammering. This limit is various in pieces of brafs 
of different thicknefs, and likewife of different quality* 
7thly, Though there are feme pieces of brafs which have 
not the property of being rendered magnetic by hammering; 
yet all the pieces of magnetic brafs, that I have tried, lofe their 
magnetifm by being made red-hot, excepting indeed when fome 
piece of iron is concealed in them, which fometimes occurs; 
but in this cafe, the piece of brafs, after having been made red-hot 
and cooled,, will attra<ft the needle more forcibly with one part 
of its furface than with the reft of it ; and hence, by turning 
the piece of brafs about, and prefenting every part of it fuc- 
ceflively to the fufpended magnetic needle, one may eafily 
difeover in what part of it the iron is lodged. 
From thofe obfervations it follows, that when brafs is to be 
ufed for the conftru&ion of inftruments wherein a magnetic 
needle is concerned, as dipping needles, variation compafles, . 
&c, the brafs ftiould be either left quite fof^ or it ftiould be 
4 chofen • 
