4IO /. H. L. VOGT 



the gabbros and norites. The difference, however, cannot be very 

 important. 



The syenitic rocks. — ^According to precision investigations at 

 the Geophysical Laboratory in Washington, the melting-point of 

 pure Ab (Na AlSijOs) or, to be quite exact, of 98 Ab+2 An, is 

 iioo°=bio°. Dr. Arthur L. Day of Washington, recently wrote 

 me that, for pure Or (KAlSiaOg) we will have to count on about 

 1200°, "di full hundred degrees higher than that of albite,"^ but that 

 precision-determinations are extremely difficult owing to the high 

 viscosity. 



The Norwegian mineralogist, Th. Vogt, in a speech in Christiania 

 Videnskapsselskap, April 16, 1920 — after the first section of this 

 work had already been sent to Chicago for publication — analytically 

 defined the binary system Or : Ab more closely,^ and further 

 estabhshed the fact that we here have to deal with a mix-crystal 

 eutectic system. The eutectic is situated at about 40 Or : 60 Ab 

 (or perhaps 35-40 Or : 65-60 Ab), without possibility of giving the 

 figures of the binary eutectic quite exactly. 



According to some investigations made in 191 2 by E. Dittler,^ 

 which, however, are only approximatives and not precision determi- 

 nations, mixtures of Or and Ab show a lowering of the melting- 

 point, which is greatest, namely almost 50°, at about 45-40 Or : 60- 

 55 Ab, that is, close to the binary eutectic between Or and Ab as 

 determined by the analytic-petrographic working method. 



My young friend Olaf Andersen (from Christiania), formerly 

 in the Geophysical Laboratory of Washington, told me that he 

 proved, by investigations made in Washington but never pub- 

 lished, that there is a lowering of the melting point for average 



'In my explanation in this Journal, 1921, pp. 335-39, I assumed that Or had 

 the same melting-point (1100°) as Ab. This is not quite a correct assumption. How- 

 ever, this has only a quite insignificant influence on the given view in the section just 

 mentioned. (Written in the Winter, 1921-22.) 



= My earlier analytic determination (in my treatise in Tscherm. Min. Fetr. Mitt., 

 24 [1905], about 40 (or 40-44) Or: 60 (or 60-65) Ab or Ab-|-An, do not count for 

 the binary system but represent the eutectic boundary curve between Or and Ab+An 

 at the presence of 10-25 per cent quartz etc. 



5 Tscherm. Min. Fetr. Mitt., 31, 1912. 



