410 Andersen — System Anorthite-Forsterite-Silica. 



Table 1. 

 Heating curves. 



Composition 





in weight per cent. 



Temperature of heat change. 



CaAl 2 Si 2 8 



MgSi0 3 





20 



80 



1259° 



30 



70 



1260° 



35 



65 



1260° 



40 



60 



1260° 



45 



55 



1259° 



50 



50 



1259° 



60 



40 



1260° 



70 



30 



1261° 



80 



20 



1260° 



seen that in all the mixtures the heating curve shows a break 

 at about 1260° C. If no further examination had been made, 

 this heat change would have been taken as a proof of eutectic 

 melting at 1260° C. The quenching experiments, however, 

 have proved that the break is not due to an ordinary eutectic 

 melting, but, as will be explained further on, to a simultaneous 

 melting of CaAl 2 Si 2 8 and decomposition of MgSi0 3 with the 

 formation of Mg 2 Si0 4 and melt. 



In some of the mixtures the heating curves showed indi- 

 cations of breaks at higher temperatures, probably corresponding 

 to points on the liquidus. These breaks were very indistinct, 

 however, and in general no other heat change than the one 

 recorded at 1260° C. was observed. 



Quenc flings. — In order to determine the compositions and 

 temperatures of the invariant points, the extension of the fields 

 of the different solid phases and the shape of the liquidus sur- 

 face a number of quenchings were made. The quenching 

 method has been described several times in papers from this 

 laboratory.* The procedure is to hold a charge of a certain 

 composition at a constant temperature long enough to assume 

 a condition ef equilibrium throughout the charge, then to cool 

 suddenly to the temperature of the air and to examine the 

 product under the microscope. By such a sudden cooling, gen- 

 erally accomplished by dropping the charge from the constant 

 temperature of the furnace into cold mercury, most silicate 

 liquids undercool readily, and form a clear glass which at ordi- 

 nary temperature remains unchanged for an indefinite length 

 of time. The solid phases that may be present in equilibrium 

 * E. g., E. S. Shepherd and G. A. Rankin, this Journal (4), xxviii, 308, 1909. 



