JUN % 1914 



AMERICAN JOURNAL OF SCIENCE 



[FOURTH SERIES.] 



Art. XL. — The Binary System MgO-SiO^ ; by N. L. Bowen 

 and Olaf Andersen. 



Introduction. 



In the course of work on two supposed binary systems of 

 which MgSi0 3 was one component, the writers have found 

 that MgSi0 3 is unstable at its melting point and can itself be 

 treated only as part of a binary system and not as a separate 

 component. The supposed binary systems mentioned must 

 therefore be treated as ternary systems and before proceed- 

 ing to the study of our separate ternary systems we have 

 worked out jointly the binary system MgO-Si0 2 which is 

 common to both. 



In compositions towards the magnesia end the temperatures 

 are beyond the range of accurate determination and towards 

 the silica end viscosity is so great as to preclude great 

 accuracy, so that only approximate results are possible, but in 

 the intermediate mixtures the determination of compounds and 

 the fixing of the invariant points have been accomplished with 

 precision. 



The Method of Working. 



The mixtures used were made up from selected quartz, 

 ground and treated with HC1, and magnesia obtained by ignit- 

 ing the precipitated carbonate to constant weight. 



In studying equilibrium in the system we have depended 

 almost entirely on quenching experiments. The method 

 has frequently been described in publications* from this 

 laboratory. It consists in holding a small charge of a mixture 

 of known composition at a measured temperature, in a platinum 

 resistance furnace, until equilibrium is attained, then cooling 

 *e. g., Shepherd & Rankin, this Journal (4), xxviii, 308, 1909. 



Am. Jour. Sci.— Fourth Series, Vol. XXXVII, No. 222.— June, 1914. 

 34 



