[ 657 D 
bars weighed 2 pounds and 6 pennyweight Troy-, 
the other 4 pennyweight lefs than 2 pounds ; and 
either of them readyly lifted with one of its ends 
better than 3 pounds and a half. 
Thefe bars were then laid down on a table, fo as 
to be nearly in one and the fame flrait line, the 
north pole of the one being next to the fouth pole 
of the other, and at the diftance of about an inch 
from it : that is to fay, that the north poles of both 
bars were pointed the fame way, but without any 
regard to the pofition of the natural meridian. 
Mr. Knight then produced a piece of natural mag- 
net, which was one of the fame he had formerly made 
ufe of, in fome experiments he had before {hewed to 
the Royal Society . This piece was in length an 
inch and in breadth A., and in thicknefs about 
*s~ of an inch at a medium, being confiderubly 
thicker at the one end than at the other. 
This piece of magnet was then applied, fo as to 
lie between the 2 firft mentioned bars, with its thin 
end clofe to the north pole of one of them, and its 
thick end clofe to the fouth pole of the other. 
After it had lain in this pofition a few moments, it 
was taken out, and upon prefenting it to the mag- 
netic needle of a (mail compafs box, it was ob (erred 
that its thinner end, the fame which had juft been 
contiguous to the north pole of one of the bars, at- 
tracted the north, end of the needle 5 and that the 
thicker end, the fame which had been contiguous to 
fouth pole of the other bar, attrafted the fouth end 
of the fame needle. 
This fame piece of (tone was then again put in 
between the bars, but in a contrary pofition 5 the 
thicker end now lying next to the north pole of one of 
Q^q q q 2 the 
