OF FERRUGINOUS BODIES. 259 
end was perfectly neutral; and that the cold part 
was a perfect magnet, having one of its poles at 
the extreme cold end of the bar, and the other 
pole distinctly marked where the metal began to 
be red hot. 
108. As the metal cooled the latter pole ad- 
vanced on the heated part, and eventually assumed 
a diffused character, similar to that shown by 
that end of soft iron which is attached to the pole 
of any bar magnet. . 
109. My next experiment was on a bar magnet 
about sixteen inches and a half long, three quar- 
ters of an inch broad, and half an inch thick. 
This bar was heated in the middle part, whilst the 
two extremities, each about one-third of the whole 
length, were kept as cool as possible by frequent 
applications of cold water. When the central 
part had arrived at a bright redness, the bar was 
taken from the fire and placed on two bricks, one 
under each end. Thus situated, it was examined 
by a compass-needle, and found to be converted 
into two distinct magnets, one at each end or cool 
part, and separated from each other by the central 
heated part, which was perfectly neutral. 
