Le S 
_ June 23, 1923] 
NATURE 
861 

~ 
The Constitution of the Alloys of Iron and Nickel. 
TBE joint pas by Dr. Hanson, of the National 
Physical Laboratory, and Mr. J. R. Freeman, 
jun., of the United States Bureau of Standards, on 
the above subject, presented at the May meeting 
of the Iron and Steel Institute, is one of great value. 
Of the elements used for alloying with iron for 
scientific and industrial purposes, none is more 
important than nickel. Moreover, the fact that it 
is an invariable constituent of meteoric iron, which 
may be regarded as a natural nickel-iron alloy, has 
invested the question of the equilibrium of these 
two elements with a high degree of interest. Its deter- 
mination, however, has presented serious problems, 
due, on ove hand, to the high melting ranges of 
the alloys and consequent pyrometric difficulties, 
and, on the other hand, to the difficulty of preparing 
the alloys free from carbon. Three years ago Dr. 
and Mrs. Hanson carried out a revision of the con- 
stitution of these alloys at 
temperatures below go00° C. 
The present publication 
completes this work right 
up to the liquidus of the 
system. 
The alloys were prepared 
by melting the purest 
materials available in a 
“carbon ring’’ furnace; 
50-gram melts were made 
in crucibles of fused alu- 
mina enclosed in a refrac- 
tory muffle made of a 
mixture of China clay with 
Io per cent. of alumina. 
Purified nitrogen was passed 
into the muffle to prevent 
oxidation of the melt. In 
taking cooling: curves the 
melts were inoculated by means of a sawdust of 
iron or nickel for the pure metals or a mixture of 
the two for the alloys, which prevented supercooling. 
Platinum rhodium thermo-couples suitably pro- 
tected were used for determining the temperatures, 
but they rapidly deteriorated with use, and in the 
case of the alloys rich in iron, possessing the highest 
melting points of the series, it was necessary to 
discard from one to two inches of the wire at the 
bulb end after each determination and remake the 
junction. The authors succeeded in keeping errors 
in temperature measurement, due to couple con- 
tamination, within 2°. 
The results are shown in diagrammatic form in Fig. 1. 
It will be seen that, starting from pure iron, the 
liquidus curve drops from 1530° to 1502° at a con- 
centration of about 5°8 per cent. of nickel. Over 
this range 6 iron separates. The solidus is the line 
AEB. Within this triangle liquid and crystals 
co-exist. The area AEG represents the limits of 
existence of 6 iron. Below 6 iron inverts to y. 
It is somewhat remarkable that this transformation 
is ay raised by the addition of nickel, as shown 
by the line GE, and 3 per cent. of this element is 
sufficient to raise the transformation point by 100°. 
The line EFB se haar the equilibrium between 
6 iron, y iron and liquid, and its meaning may be 
expressed in the following ties :—Solid E (6 iron) + 
liquid (8) react to form solid F (y iron). The exact 
oma of the point F has not yet been determined. 
he experiments thus indicate that the maximum 
solubility of nickel in 4 iron is 3 per cent, and that 
this occurs at a be i een of about 1500°. 
To the right of B there is an entirely different 
NO. 2799, VOL. 111] 
° 19 20 
NICKEL - PER cent 
state of affairs. The liquidus curve drops con- 
tinuously and much more slowly than hitherto from 
B to a minimum at C (1436° C.), the trough being 
very shallow, and then rises continuously to D, the 
melting point of pure nickel (1452° C.). Over the 
whole of this range there is complete miscibility 
between the metals in the solid state. Nickel, 
therefore, is much more soluble in y than in 4 iron, 
and the alloys consist when just solid of a continuous 
series of solid solutions. The solidus between 
B and D has, however, not been determined by 
experiménts and is shown dotted. The authors 
state, however, that it is very close to the liquidus 
line, and throughout the range from B to D, all 
the alloys have very sharp freezing points similar 
to those of pure substances. Clearly, therefore, 
the temperature range between the liquidus and 
solidus is very small. 
IRON-NICKEL ALLOYS 

Fic. rf. 
It is always difficult to determine the exact position 
of the minimum point of two curves of shallow 
inclination, and between 65 and 70 per cent the 
freezing point curve was found to be approximately 
constant, so that the exact location of C cannot 
be stated nearer than between these limits. Within 
the same range a second and smaller arrest point 
was observed between 2° and 5° below the liquidus. 
These are shown in the diagram. The authors have 
carefully tested whether these indicate the existence 
of a eutectic, but with negative results, and their 
conclusion is that C is the composition corresponding 
to a minimum in the freezing point of a continuous 
series of solid solutions. 
In the latter half of the paper an account is 
given of attempts made to differentiate by means 
of the microscope between 6 and y iron, and to 
establish the fact that the change from one to 
the other is accompanied by a recrystallisation of 
the material. Transformations in iron and iron- 
nickel alloys take place so rapidly on cooling, that 
it is impossible to preserve by quenching the 6 and + 
modifications. Accordingly attempts were made to 
stereotype the structures existing at high tempera- 
tures by a vacuum etching of polished surfaces 
previously prepared. No clear indications, however, 
were obtained, and the surface markings on the 
specimens after this treatment were very complex. 
Experiments were carried out on polished specimens 
placed in narrow silica tubes filled with nitrogen 
at 20 mm. pressure, which were inserted in a furnace 
at 1300° C. Two tubes were used. After two hours 
at the above temperature, one of them was removed 
and cooled quickly, the other was raised rapidly 
