156 RePORT—1890. 
designates a iron; between 750° and 850° the change is in process, and 
above 850° C. he asserts that iron enters the other molecular state, which 
he designates 8 iron. It is, then, to the latent heat of allotropy that 
Osmond attributes the recalescence observed in iron, heat being absorbed 
to produce this change at the critical point during heating, and liberated 
during cooling at a somewhat lower temperature. Iron, according to 
this hypothesis, is a polymorphous element like sulphur, phosphorus, 
&c. Sudden cooling from a white heat, when the change into f iron 
has occurred, should tend to preserve the iron in this allotropic state ; 
but this is not the case, except to a small extent, and hence Osmond 
maintains that it is the presence of carbon in the iron which keeps the 
iron in the £ condition when suddenly cooled. Hardened steel would 
thus owe its properties principally to the presence of f iron, which is 
hard and brittle at ordinary temperature: ‘both the iron and the carbon 
in hardened steel preserving more or less completely in the cold the con- 
dition which they possessed at a high temperature.’ 
We think, however, that the evidence adduced by Osmond on behalf 
of his theory of recalescence in iron is as yet insufficient. No doubt iron 
does exist in an allotropic modification at a high temperature, but the 
electrolytic iron with which Osmond experimented contained 0:08 per 
cent. of carbon, very nearly as much as some of the pure steels with 
which we have experimented, which contained 0:1 per cent. of carbon. 
It is to the influence of this small amount of carbon present in Osmond’s 
electrolytic iron that we are inclined to attribute the feeble recalescence 
which he observed in his specimen. The effect, (a) of this residual 
carbon, and (b) of the mechanical treatment the specimen has received, 
such as hammering and wire-drawing, have yet to be investigated, and 
this we hope to undertake during the next year. If it be possible to 
keep iron in the B condition when cold, it should not only be hard and 
brittle but non-magnetic, and this has not yet been proved. We have 
made some experiments on this point by suddenly quenching at a white 
heat fine iron wires in cold mercury, and here will merely state that 
their magnetic susceptibility was not destroyed. Manganese steel, it is 
true, is practically non-magnetic, and this Osmond attributes to the part 
played by manganese in fixing the iron in 'the 8 condition, and Hopkin- 
son has shown that whatever slight magnetic susceptibility is found in 
manganese steel could be accounted for by a few little bits of pure iron 
distributed through the mass.! 
We believe that the difference in the temperature position of recalescence 
(and also of the jerk) on the up and on the down side of the curve of heat- 
ing or cooling is analogous to what is found in the heating of water. In 
a clean vessel water may be raised above the boiling-point; suddenly at 
some one point steam is formed, and the whole rapidly passes into steam, 
the change of state being accompanied by a full of temperature and large 
absorption of heat. Similarly steam in cooling down may be lowered 
below the normal point of condensation, when from some cause, such as 
the presence of solid particles, condensation begins and rapidly proceeds, 
accompanied by a rise of temperature.? The retardation of this change 
Mr. Tomlinson ’ considers to be due to the great internal friction which 
exists in iron at a red heat; in consequence of this the change takes place 
1 Phil. Trans., April 1885. 2 Newall, Camb. Phil. Soc., Jan. 1888. 
8 Phil. Mag., Feb. 1888. 
