Diopside — Forsterite — Silica. 261 



an assumption is illustrated in the magnesia-rich mixtures of 

 the present investigation and especially in the simple case of 

 the mixture of composition MgSi0 3 . The behavior of a 

 liquid of this composition, when cooled, illustrates the fact 

 that, in the liquid state, the whole cannot consist simply of the 

 compound MgSi0 3 . All possible compounds must be con- 

 sidered ; that which separates in the solid form will be the 

 one which, under the conditions, happens to be the first to 

 reach its limit of solubility. 



In the case of liquid of composition MgSi0 3 we must con- 

 sider that, among others, Mg 8 Si0 4 molecules are present and 

 that, on cooling, the liquid becomes saturated with these at 

 1577°, forsterite separating out. On further cooling forsterite 

 continues to separate, the liquid always remaining at the point 

 of saturation with respect to that substance. At 1557° the 

 liquid becomes saturated with respect to MgSiG 3 and clino- 

 enstatite begins to separate. In the liquid an equilibrium 

 exists which can be represented by the equation 



2MgSi0 3 ^z± Mg 2 Si0 4 + Si0 2 



and at any temperature there is a definite equilibrium constant, 

 that is, the relative proportions of the three compounds are 

 fixed. When, at 1557°, MgSi0 3 separates from the liquid it 

 is necessary that this reaction should proceed in the direction 

 from right to left in order to maintain these proportions. 

 This reaction lessens, however, the concentration of Mg 2 Si0 4 

 in the liquid and since the liquid must remain at the point of 

 saturation with respect to forsterite, forsterite must dissolve. 

 This cycle continues, forsterite dissolving and clino-enstatite 

 precipitating, until the whole consists of crystalline clino- 

 enstatite. 



In all the compositions lying along the line between MgSi0 3 

 and CaMgSi 2 6 as far as 75 per cent CaMgSi 2 6 we have, in 

 the separation of forsterite, this direct evidence of the same 

 equilibrium reaction in the liquid. It should not be assumed, 

 because forsterite does not separate from liquid CaMgSi 2 6 

 and the mixtures close to it, that this reaction does not occur. 

 On the contrary, it simply happens that liquid OaMgSi 2 6 

 becomes saturated on cooling with diopside first, and for this 

 reason the whole crystallizes to diopside. Among others, a 

 possible equilibrium in liquid diopside could be written 



2CaMgSi 2 6 t-> 2CaSi0 3 + Mg 2 Si0 4 4- Si0 2 . 



Crystallization of diopside simply causes the reaction to pro- 

 ceed from right to left and the whole becomes solid diopside, 

 the crystalline product, by itself, giving no clue to the state 

 of combination in the liquid. 



