78 
MR. S. W. J. SMITH ON THE THERMOMAGNETIC ANALYSIS OF 
Again, if it cannot be said that the chemical evidence disproves the view that the 
eutectic contains 40 per cent, of nickel, it can at least be said that there is as much in 
favour of the view that it contains between 25 and 30 per cent. 
It would be unwise to generalise from the thermomagnetic analysis of a single 
specimen ; but, as will be seen below, the Sacramento meteorite certainly possesses a 
constituent containing between 25 and 30 per cent, of nickel, of which the internal 
structure is complex and, as seen above, a consideration of the chemical data in detail 
distinctly supports the view that there is a constituent in all octahedral meteorites 
which contains between 25 and 30 per cent, of nickel and which consists of 
components unequally attacked by acids. 
Finally, the presence of a 25 to 30 per cent. Ni constituent can be explained 
naturally, on the assumption that it is a eutectic, by means of the already known 
thermomagnetic properties of nickel-iron alloys ; while the existence of a 40 per cent. 
Ni eutectic requires, for its explanation from the experimental curves, hypothetical 
extrapolation. 
Section VI.— -An Interpretation of the “Irreversibility” of Nickel-iron 
Alloys. 
§ 1. From the point of view of the theory of solution, considerable interest attaches 
to those experiments of the present paper in which the nature of the thermal 
hysteresis of the nickel-iron alloys is examined. 
The qualitative resemblance between the thermal hysteresis in the return of 
magnetism in nickel-iron alloys and the phenomena of supersaturation has already 
been commented upon {of., e.g ., Guillaume, ‘ Recherches sur le Nickel et ses 
Alliages,’ 1898). 
A comparison of the properties of nickel-steel with those properties of super¬ 
saturated solutions recently described by Miers (‘ Nature,’ August 24, 1905, vol. 72, 
p. 412) suggests that the analogy is even closer than has appeared hitherto. Miers 
finds that a solution of a salt of given strength, saturated at a temperature d l , 
remains supersaturated, when the temperature is lowered, until a definite temperature 
0 2 is reached. Throughout the range 0 ] to d 2 , crystallisation will not begin, however 
long the solution be allowed to stand (if surface evaporation is prevented), unless 
crystals of the salt are added from without. The value of 0 2 , like that of 6 1} depends 
upon the strength of the solution. If crystals are introduced at temperatures 
between 6 1 and d 2 , they grow at the expense of the salt in solution ; but the growth 
proceeds slowly even if, in order to facilitate uniform distribution of the salt, the 
solution is kept stirred. 
When, however, the temperature reaches 0 2 , a change in the condition of equilibrium 
of the solution occurs, which is accompanied by spontaneous generation of crystalline 
nuclei. Tbe solution, saturated at 0 U may be said to pass from the metastable to the 
