THE BEHAVIOR OF INCLUSIONS IN IGNEOUS MAGMAS 533 



converted into clino-enstatite. Effectively, then, the liquid (M) is 

 supersaturated with forsterite. It cannot dissolve forsterite but 

 can only convert it into clino-enstatite, the phase with which it is 

 saturated. This principle is capable of general application and 

 we may state that a hquid saturated with any member of a discon- 

 tinuous reaction series is effectively supersaturated with all higher 

 members of the series; it cannot dissolve them but can only con- 

 vert them into the phase with which it is saturated. 



In connection with the specific case we have discussed, it should 

 be noted that the amount of clino-enstatite formed is not simply the 



n^^LO^ zo /i^SlOj ^0 



Fig. 7. — Lower temperature portion of the equilibrium diagram, forsterite-silica 



chemical equivalent of the forsterite changed. The chemical equiv- 

 alent would be somewhat less than one and one-half times the for- 

 sterite, whereas the amount of clino-enstatite formed is much greater, 

 being in fact about five times the amount of forsterite. In other 

 words, the action is not simply an addition of silica to the forsterite, 

 with consequent impoverishment of the liquid in silica, for the liquid 

 cannot have silica subtracted from it without passing under the 

 clino-enstatite saturation curve, that is, without precipitating cHno- 

 enstatite. In order to convert the forsterite inclusions into clinoen- 

 statite, the liquid must precipitate a large amount of clino-enstatite 

 from its own substance, and the action uses up a large amount 

 of the liquid. The liquid left has, however, in this binary case, the 

 same composition as the initial liquid, if the temperature is kept 



