A COOLING MIXTURE 173 



sium metasilicate mentioned above, and suppose it to have been melted 

 and thorotighl}^ mixed in the proportion of 80 parts CaSiOg to 20 parts 

 MgSiOg (figure 7), and then to be gradually cooled preparatory to ob- 

 serving solidification, we shall encounter not one change of state only 

 during the cooling, but three, each of which is definitely characterized by 

 a change in the energy content of the system, and is therefore a physical 

 discontinuity which completely defines the stability of the substance with 

 respect to temperature, provided only we are in position to make an in- 

 telligent interpretation of what occurs. To follow this case, the first 

 release of heat on cooling appears at about 1405 degrees, and represents 

 the temperature at which, if no disturbance of the equilibrium takes 

 place, the excess of calcium metasilicate begins to crystallize out. At 

 about 1350 degrees, calcium metasilicate and diopside will crystallize in 

 eutectic mixture with a second release of heat. From this point down- 

 ward all of the solution has become a solid mixture of these two ingre- 

 dients, but at about 1190 degrees, if the mechanical forces which oppose 

 any rearrangement of the molecules be not too great, the calcium meta- 

 silicate will go over from the crystal form known as pseudo-wollastonite 

 into true wollastonite with a heat change similar to the two preceding, 

 but smaller in magnitude. The melting of this mixture is exactly the 

 reverse process. If the original mixture had contained only 10 per cent 

 of magnesium metasilicate instead of 20 per cent, all the other conditions 

 remaining the same, crystallization would have begun when the solution 

 reached a temperature of 1465 degrees instead of 1410 degrees, the other 

 two changes following as before. With a still smaller percentage of mag- 

 nesium metasilicate, the first crystals might have appeared as high as 

 1500 degrees. 



The same two ingredients in the proportions CaSiOg 47, MgSiO,^ 53, 

 would have given but a single change of state (the melting or crA^stalliza- 

 tion of pure diopside), while with CaSiOg 20, MgSiOg 80, three changes 

 of state would have occurred as in the first instance, except that the excess 

 component which first appears is now a-magnesium silicate instead of the 

 corresponding lime compound, and the inversion which follows is an 

 enantriopic change in the magnesium silicate. 



If the above explanation has been clear, it offers an illustration not 

 only of the crystallization of an excess component, but of the eutectic and 

 a subsequent inversion of one of the components of the eutectic. It illus- 

 trates also the continued lowering of the melting temperature of one 

 mineral (CaSiOs) through the addition of increasing quantities of an- 

 other mineral. This concrete illustration is quite typical of the general 

 case, in which the solution of one silicate in another will lower its melting 



XTTI— Bull. Geol. Soc. Am.. Vol. 21, 1000 



