April 13, 1899] 



NA TURE 



567 



is placed in the piece of metal n, and the other in a compen- 

 sating piece of copper platinum of fire-clay c. A sensitive 

 galvanometer Go, connected to both thermo-couples, measures 

 on a large scale the difference between the temperatures of H 

 and c, and magnified records of the evolutions of heat in B can 

 thus be obtained, which are not affected by the general fall of 

 temperature of the system The actual temperature of the piece 

 of metal B is simultaneously registered by the less sensitive gal- 

 vanometer G, in the usual way. In the new method, therefore, 

 the heat lost by the cooling mass of metal B (Fig. 2) is com- 



pensated or balanced by the heat lost by a mass of platinum c. 

 The result is, that the effect on the galvanometer Gj of any 

 evolution of heat in the cooling mass B is greatly augmented. 

 The heat suddenly evolved by the mass of iron or steel B, which 

 is liable to molecular change, is, as has already been indicated, 

 not masked by the fact that the mass is itself rapidly losing heat, 

 as the temperature of the entire system does not affect the 

 sensitive galvanometer, and the heat which is evolved by the 

 mass E is free to make itself felt. Hence the curves recorded 

 by its mirror possess extraordinary sensitiveness. In the Figs. I 



was deposited from a solution of ferrous chloride which had 

 been purified with scrupulous care, the anode being a plate of 

 electro-iron. The iron so deposited weighed five grammes, and 

 its appearance, magnified four diameters, is shown in the accom- 

 panying Fig. 3. 



It was then arranged as shown in Fig. 2, and placed in a 

 porcelain tube glazed inside and out and rendered vacuous by 

 the aid of a mercurial pump which also enabled the gas evolved 

 from the iron to be collected. Hydrogen was freely evolved as 

 the portion of the tube containing the iron was gradually heated ; 

 but, although the evolution of gas never 

 absolutely ceased, the amount of hydrogen 

 delivered by the mercurial pump with 

 which the porcelain tube was connected 

 was very small when the iron attained a 

 temperature of some 1300° C. A cooUng 

 curve of this iron after four successive 

 heatings is shown in Fig. 4, on J of the 

 actual scale it was recorded, and it will at 

 once be evident that in this curve at least 

 three hitherto unobserved points are re- 

 vealed. These points occur respectively at 

 580°, 487°, and 261°. The coordinates are, 

 as usual, time and temperature, but the 

 temperature represents on a large scale, 

 molecular evolutions of heat, and not the 

 temperature of the mass under examination. 

 There is at A the point at 1130° C. Then 

 at B there is the ordinary Ar. 3 of Osmond, 

 which in this case occurs, not as in mild 

 steel at the normal temperature of 850° C. , 

 but at 895° C. When the mass continues 

 to cool down there is, as it was anticipated 

 there would be, the point Ar. 2, which in 

 this case occurs at 770^ C. The carbon 

 point Ar. i could not be expected to occur 

 in iron of so high a degree of purity, and it 

 does not exist : but there is evidence of 

 evolution of heat at a point which is 

 between 550° and 600' C. It is difficult 

 to fix this point accurately, it seems to 

 vary somewhat in successive curves. The 

 next point, at which heat evolved, is a new one of extra- 

 ordinary interest. It occurs between 450' and 500° C, and 

 there is evidence to show that it is connected with the retention 

 of hydrogen by the mass of iron, even though it had been 

 heated to 1300' C. Finally there is a small point at 270° C. , 

 that is, at a temperature of no less than 400^ C. below redness. 



The significance of these new points which, seem to be con- 

 nected with the retention of hydrogen by the iron, though they 

 may be true iron points, revealed by the presence of hydrogen, 

 is to some extent considered in the Report. All that need be 



-Photographic Record (i actual size) of the Cooling of Electro-Iri 



and 2 the arrows show the direction of the currents. Those 

 with feathers indicate the direction of the current which is due 

 to the difference of temperature. This difference is caused by 

 the excess of heat in the iron B, as compared with the platinum 

 C. The featherless arrows show the direction of the current 

 through the unsensitive galvanometer Gi, which records ordinary 

 cooling curves. 



Reference to a special case, furnished by the cooling of elec- 

 tro-iron from a white heat, will serve to make this clear. 

 Electro-iron was deposited on a thermo-junction protruding 

 from a glass tube into which the wires were fused. The iron 



pointed out here is, that it is important to have obtained a new 

 and delicate method of recording molecular changes which take 

 place in metals and alloys as they cool down from the fluid 

 state to the ordinary temperature. 



By the aid of a long series of such curves Sir William Roberts- 

 Austen has laboriously investigated the carburised-iron series of 

 alloys, which are usually known as steel and cast iron. He 

 shows that they behave exactly as certain saline solutions do, 

 and he has thits afforded a basis for much of the hitherto some- 

 what obscure procedure in the industrial treatment of iron 

 and steel. 



NO. 1537, VOL. 59] 



