146 REPORT—1 890. 
Professor J. A. Ewing has shown recently,! in his beautiful experi- 
mental model representing molecular magnets, how a state of magnetic 
instability may occur in the magnetic metals as a certain critical tempera- 
ture is approached, the chief facts of permeability and retentiveness, and 
what Ewing terms hysteresis, being explicable by supposing that a mag- 
netised bar is made up, as in Weber’s hypothesis, of molecular magnets, 
but ‘constrained by no other forces than those due to their own mutual 
attractions and repulsions;’ increase of permeability due to rise of tem- 
perature, for magnetising forces far short of saturation, being caused by 
the expansion and separation of molecular centres creating a reduction of 
stability. And as regards the sudden loss of susceptibility at the critical 
temperature, Ewing conjectures that the violence of the oscillation of 
the molecular magnets at this temperature may cause a state of rotation 
to be developed, wherein, of course, all magnetic polarity would disappear. 
Professor Ewing’s suggestive paper shows us that we may well expect 
other remarkable phenomena,—and abrupt changes in the physical proper- 
ties of the magnetic metals are found to take place,—at this critical 
temperature. ‘I'he Committee have been engaged in investigating some 
of these. 
I. We will first take the sudden anomalous expansion observed when 
steel and some specimens of iron wire cool from a white heat first noticed 
by Gore in 1870, and the corresponding anomalous contraction on heating 
first noticed by one of us in 1873.2, The observation of these effects is 
extremely easy. It is only necessary rigidly to fix one end of an iron or 
steel wire and attach the other end to a multiplying lever, or observe 
through a reading microscope a mark on the free end, when on heating 
the wire either by a gas flame or an electric current the following 
phenomena are observed. The wire steadily expands as the temperature 
rises till a low red heat is reached, when a halt occurs, then a sudden 
momentary retraction of the wire takes place, after which expansion 
continues to the fusing point. On cooling, the wire regularly contracts 
till a temperature a little lower than that at the jerk on heating is reached, 
when a sudden momentary elongation of the wire occurs, and then con- 
traction ensues till it is cold. If the wire be vertical or horizontal, 
with or without tension, the effect is equally present. To perceive the 
jerk on cooling, it is, however, absolutely essential that the temperature 
of the wire should be raised above that at which the jerk on heating 
occurs, otherwise no anomalous effect is observed. In the experiments 
made by one of us in 1875 but not hitherto published, Barrett found that 
in some specimens of steel wire two anomalous contractions on heating 
and expansions on cooling were noticed, the feebler one taking place at a 
lower temperature. Further, that in some specimens of iron 20 anomalous 
expansion or contraction was noticeable, whilst more generally in other 
svecimens the effect on cooling only was noticed, and that usually this 
effect could be wiped out by a few successive heatings and coolings. But 
in steel the jerk in cooling was always present and was not wiped out, 
though at first slightly reduced, by repeated incandescence and cooling. 
Further, judging by the amount of the expansion of the wire, the effect 
on heating took place, as stated above, at a slightly higher temperature 
than the jerk on cooling. 
A series of unpublished experiments were long since made by the 
1 Phil. Mag., Sept. 1890. ? Barrett, Phil. Mag., Jan. 1874. 
