586 N. L. Bowen— Melting Phenomena of the 



refractive index. By trying various mixtures one was readily 

 found which exactly matched the glass, the comparison being 

 made by the Becke line method. The index of the liquid was 

 then immediately determined on the refractometer. The 

 indices of the pure, artificial glasses of the albite-anorthite 

 series had previously been determined at this laboratory by 

 Larsen,* so that when the index was known the composition 

 was also known. 



The results obtained are tabulated below in such form as to 

 indicate the extent of agreement of the two methods. 





Table III. 





Composition 



Composition 



Composition 



of charge Initial 



Measured Index from index 



from liquidus 



Mol. % An. Condition. 



Temperature. of glass. Mol. % An. 



Mol. % An. 



66-6 (A^AnJ Glass 



1482° 1-542 63 



63 



Glass 



1462 >l-534<l-538 53'5-58 



54-5 



Glass 



1454 1-532 51 



51-5 



50-0 (AbjAnJ Cryst. 



1405 1-520 37 



36 



Glass 



1385 >1-514<1-518 30 — 35 



31 



Glass 



1432 1-525 43 



44 



33-3 (Ab 3 AnJ Cryst. 1324 1-506 20 19 



By means of fig. 2 the composition of the liquid phase for 

 each experiment, determined from its refractive index, may be 

 quickly compared with the composition determined by the 

 temperature method of fixing the liquidus. The agreement is 

 excellent. 



The temperature method is somewhat more accurate and much 

 more generally applicable than the refractive index method. 

 The latter, however, provides entirely independent confirma- 

 tion of the former, and . serves to demonstrate beyond doubt 

 the great difference in composition between liquid and solid 

 phases at any temperature. For example, a mixture of total 

 composition A^Ar^ at 1385° gives a glass nearly 20 per 

 cent richer in albite, and since about one-half of the charge 

 is crystalline, it follows that the crystals must be approximately 

 20 per cent richer in anorthite. The actual difference between 

 solidus and liquidus at 1385° is seen from fig. 1 to be 38 per 

 cent. 



From the zoned feldspars of nature it has long been known 

 to geologists that a very considerable difference of this kind 

 must exist. Vogtf has even ventured an estimate, for which 

 he makes no claim of exactness, of the magnitude of this differ- 

 ence based on a comparison of the composition of the total 



* This Journal, xxviii, p. 283, 1909. fT.M.P.M., xxiv, p. 514, 1905. 



