99 
of Edinburgh, Session 1872-73. 
Curve T. of each day’s experiment represents the rate of diminu- 
tion of the-hot pole in terms of its temperature; Curve II. that of 
the cold pole in terms of the temperature of the hot pole ; and 
Curve III. shows how the magnetic strength of the one pole 
changes relatively to that of the other. 
Perhaps the most important fact made out from these experi- 
ments was that already mentioned, viz., that whatever he the tem- 
peratures of the ends of the magnet, the magnetic strength is 
approximately the same in both poles. This was shown by plotting 
the deflections of the little magnets in terras of one another; for 
it was then found that the result was a straight line, which proves 
that the rate of diminution or increase of magnetic strength is the 
same in both poles ; and, therefore, if they be of the same strength 
at the commencement of each experiment, when the magnet is of 
the same temperature throughout, they will of course remain the 
same during the whole experiment, when the poles are of different 
temperatures. (See figure III. of each day’s experiment.) When 
the pole which had been heated was allowed to cool again, the line 
obtained by plotting the deflections of the two poles was still 
approximately a straight one, though not exactly coinciding with 
that obtained for it when being heated. This, however, is quite 
satisfactorily accounted for by the alteration of the zeros, which 
was never the same in both, and therefore the measured strength 
differed from the true strength more in one pole than in the other. 
In the first communication it was pointed out how the mere heat- 
ing of the pot produced in some unaccountable way an alteration 
of zero. The fact that the poles are the same strength throughout 
also accounts for the general similarity of Curves I. and II. for 
each day. 
It is to be observed that Curves I. and II. become smoother day 
after day, as if the boiling so altered the molecular constitution of 
the steel, as to enable it to conduct more easily a state of magnetic 
distribution. 
It is interesting also to notice that the rate of diminution of 
magnetic strength decreases on each successive experiment, and 
that the two branches of each curve approach nearer and nearer to 
each other, thus showing a greater unwillingness on the part of 
the magnet to lose any of its magnetism permanently with heating. 
