CONSTITUTION AND TEMPERATURE ON MAGNETIC SUSCEPTIBILITY. 
129 
second experiment, crystallization took place at a temperature 70° C. below the normal 
melting point. In each case, when the crystals were heated, fusion took place at the 
normal temperature, 48° C. The substance was melted by hot water contained in a 
Dewar tube, and the temperatures below that of the room were obtained with a tube 
which had been cooled by liquid air. 
It is apparent that with this substance a definite hysteresis effect, due to temperature, 
is obtained, similar to that which was discovered by the late Prof. Hopkinson*, in the 
case of nickel-steels. If we regard the paramagnetic state of the nickel-steel as 
isotropict, the temperature hysteresis effects in the two cases are completely analogous. 
The paramagnetic solid corresponds to the diamagnetic liquid. On cooling, the nickel- 
steel passes the critical temperature^ without crystallizing just as the benzophenone 
passes the normal fusion point and assumes a gelatinous form. Continued cooling in 
each case causes the more stable crystalline form to appear. This is accompanied by 
a large rise of the paramagnetism of the nickel-steel (ferro-magnetism) and by a fall 
of the diamagnetism of the benzophenone. If now the nickel-steel be heated, it loses 
its ferro-magnetism at the critical temperature—a temperature which is characteristic 
for a given alloy and is lower the greater the proportion of nickel in the steel—while 
the benzophenone regains its larger diamagnetic property at the normal fusion point— 
a temperature which is lower the greater the impurity present. Further, the heat 
produced (recalescence) when the ferro-magnetic state appears on crystallization of 
the nickel-steel is analogous to the heat produced (heat of formation of the crystals) 
when the benzophenone crystallizes. The molecular changes are of the same nature 
in the two cases; the only difference is in the character of the magnetic property 
possessed by the molecules and by which such molecular changes are indicated. § 
* ‘Roy. Soc. Proe.,’ vol. XLVIII., p. 1, 1890; or Ewing’s ‘Magnetic Induction in Iron and other 
Metals,’ p. 184, et seq. Hopkinson referred to the paramagnetic state above the critical temperature as 
the non-magnetizable state. 
[t At least we may regard it as an allotropic modification of the alloy. See the article by Guillaume, 
‘Rapports du Congres Int. Paris,’ 1900, vol. 1; also Dumas, ‘ Journ. of the Iron and Steel Institute,’ 68, 
p. 255, 1905. It should be mentioned that the results described in this communication concerning change 
of state are in accordance with the work of Tammann, ‘ Wied. Ann.,’ LXII., p. 285, 1897. See also the 
article by Spring, ‘Paris Reports,’ 1900, and Wpietham’s ‘Theory of Solution,’ Chap. II., where 
additional references to Tammann’s work are given .—Note added March, 191J /..] 
I It would be more correct to speak of the “range of temperature of transformation” instead of 
“ critical temperature.” 
[§ Mr. M. Owen has observed that the specific susceptibility of super-cooled gallium, between 30° C. and 
16“ C., is constant and equal to that of the liquid above the normal fusion point. The crystalline state is 
much more diamagnetic and y c remains nearly constant until the fusion point, 30° C., is reached. 
My attention was drawn to this point of priority, after my work had been completed, by the curve 
given in Owen’s paper, ‘Ann. der Phys.,’ IV., 37, p. 693 (fig. 4 f), 1912. The English translation of this 
paper, ‘Versl. Kon. Ak. v. Wetensch.,’ Amsterdam, XIV., p. 637, 1912, which first came to my notice, 
contains no diagrams and I had overlooked the brief reference there made, at the foot of p. 642, to the 
behaviour of gallium.—Note added March, 191^P\ 
VOL. CCXIV.—A. 
s 
