PRINCIPLES UNDERLYING METAMORPHIC PROCESSES 495 
rium and therefore is a mix-crystal. The solid forms actually 
known to us are not the pure pseudo-components themselves, but 
are solid phases in an equilibrium which is displaced by external 
circumstances toward one side or the other. The displacement of 
equilibrium in the solid phase produced by change of external con- 
ditions may proceed until unmixing supervenes, i.e., until an 
inversion point is reached. 
This theory has already justified itself as a working hypothesis 
at least; it has enabled Smits to co-ordinate certain phenomena and 
hence in part to predict the behavior of certain systems. This 
success indicates the great probability of the existence of internal 
equilibria, and in this respect it is in complete harmony with a 
number of other lines of evidence. It leads us to look upon a 
single crystal, not as a simple body, but as a more or less compli- 
cated configuration the behavior of which depends on the state of 
equilibrium within the crystal. 
Moreover, it possesses a great advantage as compared with the 
old point of view in that it affords a clue to the problem of the 
apparently arbitrary appearance of metastable modifications. For 
if we know the diagram of such a pseudo-unary system—that is, 
if we know the equilibrium lines in both the liquid and the solid 
phase as well as the mix-crystal and liquidus curves—we can fore- 
tell which factors will promote, or hinder, the formation of a given 
metastable product. For instance, if the speed of attainment of 
equilibrium is small and the temperature change is rapid, the 
equilibrium will not have time to adjust itself to the changing con- 
ditions; unmixing therefore will occur at a composition (with 
respect to the pseudo-components) and at a temperature which 
differ from those corresponding to continuous real equilibrium—in 
other words, the resultant products in the two cases are different. 
The position of the various equilibrium lines and the rates at which 
the various equilibria are established under the particular condi- 
tions are, in short, the factors that determine which product actually 
appears. Hitherto the only aid in the prediction of what would 
appear has been that derivable from the so-called Ostwald rule,* a 
' The rule, as stated by Ostwald, is: In all reactions the most stable state is not 
straightway reached, but the next less stable or that state which is the least stable of 
the possible states. It may be observed that this alternative mode of statement 
insures the validity of the rule in a large number of cases. 
