Plagioclase Feldspars. 579 



CaMg(Si0 3 ) o 14160 1/2 hr. cryst. Nov. 22, 1912 



u " 14200 1/2 hr. glass Nov. 22, 1912 



" " 14150 1/2 hr. cryst. Jan. 10, 1913 



" " 14180 1/2 hr. glass Jan. 10, 1913 



Melting point between 14160 and 14180 M.V. 



Value by standard element 14231 M.V. Correction + 60 M.V. 



CaAl 2 Si„0 8 16100 1/2 hr. glass ■ Nov. 22, 1912 



" 16050 1/2 hr. glass Nov. 22, 1912 



" 16010 1/2 hr. cryst. Nov. 22, 1913 



" 16030 1/2 hr. cryst. '& glass Jan. 10, 1913 



u 16000 1/2 hr. cryst. only Jan. 10, 1913 



Melting temperature 16030 M.V. 



Value by standard element 16150 M.V. Correction + 120 M.V. 



It will be noted that after several weeks continuous use 

 (Nov. 22 to Jan. 10) the thermoelement was found unaltered. 

 By interpolating between these values, the correction to be 

 applied to the individual readings, in order to express them in 

 terms of the standard element of Sosman and the nitrogen scale,* 

 is readily found. These corrections have been applied to all 

 readings in changing to degrees. 



The application of this method of quenching to the specific 

 problem of the feldspars was directed towards two definite 

 ends; first, the determination of the exact limits of the melting 

 interval for a given composition, and second, the determination 

 of the composition of the solid and liquid phases present at a 

 given temperature within the melting interval. 



Determination of the Limits of the Melting Intervals. 



The location of the temperature of the lower limit of the 

 melting interval (solidus) was accomplished by starting with 

 completely crystalline material of a certain composition and 

 holding it for one hour at successively higher temperatures 

 until a temperature was reached at which the quenched prod- 

 uct showed a trace of glass. For the upper limit (liquidus), a 

 charge wholly or partially crystalline was held at successively 

 higher temperatures, again for periods of one hour each, until 

 the quenched product showed that the last tiny crystals had dis- 

 solved in the glass. This temperature was then further checked 

 by holding a clear glass of the same composition at successively 

 lower temperatures until crystals appeared in the quenched 

 product. 



The results of these trials may be most concisely presented 

 in tabular form. 



*This Journal, xxx, p. 1, 1910. Day and Sosman, High Temperature Gas 

 Thermometry ; Carnegie Institution of Washington, Pub. 157, p. 119. 



