152 3 REPORT—1890. 
M. Ledeboer, devised by M. H. Le Chatelier, and consisting of a thermo- 
electric couple of platinum and platinum rhodium alloy, associated 
with a dead beat galvanometer, Osmond has made a careful study of the 
recalescence in iron and steel. 
Osmond finds three critical points to exist in mild steel when, during 
cooling, the temperature remains stationary for a sensible interval of 
time. These three points he designates a,, a, and a3; a, being that 
at the lowest temperature about 660° C., a, about 730° C., and a3 about 
850° C. In hard steel a, only is present, but is much more pronounced, 
and occurs somewhat higher, about 700° C. In electrolytic iren, which, 
however, contained 0:08 per cent. of carbon, a, and a3 only are present, 
occurring at about 720° C. and 860° C. respectively. The temperature 
of these critical points he finds, just as we found with the anomalous ex- 
pansion and contraction, to be higher during the heating than during the 
cooling of the same specimen. Osmond, however, has not noticed any 
sudden rise of temperature at recalescence, only a longer or shorter halt 
in the cocling. But this difference probably arises from his mode of ex- 
perimenting; one of us has pointed out the necessity of precaution in this 
respect.! We have recently repeated our experiments, using a thermo- 
couple similar to that employed by Osmond, and find that there is not 
the least difficulty in observing and measuring the sudden large increase 
of temperature that occurs during recalescence. It is only necessary to 
use somewhat fine wires for the thermo-couple, to bind them to the steel 
wire under experiment, and wrap the part round with asbestos to prevent 
too rapid cooling in the air. 
The temperature of the critical point Osmond finds, as occurred with 
the jerk on heating or cooling, to be higher on the up side of the curve 
(that is, during heating) than on the down side of the curve. Like our- 
selves, he finds the critical point higher in iron than in steel, and that on 
re-heating the same sample the point of recalescence is lowered somewhat. 
Experiments we have recently made show that after the first two or three 
heatings stable conditions appear to be reached; recalescence then occurs 
at the same temperature and to the same amount on subsequent heatings 
and coolings. This is assuming the metal to be raised to the same tem- 
perature before cooling each time; if the temperature before cooling be 
not so great Osmond finds that the critical point is raised. Our experi- 
ments point rather the other way, but they need repeating. When the 
cooling of iron or steel is slow recalescence begins at a somewhat higher 
temperature, and lasts longer than when the cooling is very rapid. 
Osmond finds that when the cooling is very rapid, by quenching the 
white-hot metal in water, recalescence is entirely absent. The steel is 
thus hardened, and Osmond concludes that the latent heat of the change 
which takes place in the metal at recalescence is still in the steel, and he 
terms it the latent heat of hardening. 
IV. The difference in the temperature at which recalescence occurs on 
the up and down side of the curve of heating, has led two of our Com- 
mittee independently to suggest the explanation of the remarkable thermo- 
electric current which is produced in iron and steel by a moving source of 
heat.? If an iron or steel wire be heated to redness at any one point of 
1 Newall, Phil Mag., June 1888. 
2 Trouton, Proc. Royal Dublin Soc., 1887. Newall, Phil. Mag., June 1888. Mx. 
Trouton has found a similar but entirely transient E.M.F., caused by a moving flame 
