DIFFERENTIATION OF IGNEOUS ROCKS 409 



only a small quantity, existing as a eutectic end-magma, at some 

 lower temperature. 



A great part of the crystallization of other peridotites with much 

 olivine (most frequently between o.io Fez Si04 • 0.90 Mga Si04 and 

 0.15 Fe2 Si04. 0.85 Mg2 5104) and a Httle more bronzite, diopside or 

 diallage, etc. (mostly 15-40 per cent) than in the first case will 

 have taken place at a high temperature, about 1400°. The same 

 is also true for bronzite rocks. 



We will now discuss the most important anchi-eutectic rocks. 



As a starting point for the gabbros and the norites, we have the 

 eutectic boundary-line determined by Bowen between Ab+An and 

 diopside, with temperatures (as before at i atmosphere pressure 

 and in water-free melts) from about 1260° for 51 diops. : 49 

 Abi An4 to about 1215° for 28 diops. : 72 Ab7An3. If we substitute 

 ordinary iron-bearing hypersthene or diallage for the iron-free 

 diopside this will result in the lowering of the above-mentioned 

 temperature but not to any specially important degree. Most 

 gabbros and norites contain a small surplus above the eutectic 

 boundary, in some cases of plagioclase and in others of a ferro- 

 magnesian silicate. Hence the crystallization wiU most frequently 

 commence at 1250° or 1275°, and the main part crystalHze along 

 a eutectic boundary at about 1225-1200°. For some gabbros and 

 norites, diabases, basalts, etc., however, we will have to assume a 

 still lower end temperature. These values, calculated from the phys- 

 ico-chemical diagrams (for water-free magma at i atmosphere pres- 

 sure) are confirmed by an experiment made by Sosman and Merwin^ 

 concerning the melting of a diabase from Granton, New York: 

 "The remelting begins at about 1150° and is practically completed 

 somewhat below 1300°. The rock flows readily at a temperature 

 about 1225° at which temperature all is fused excepting about a third 

 of the feldspar. A little feldspar still remains undissolved at 1250°, 

 but only traces remain at 1300°." 



The diorites, owing to the somewhat higher per cent Ab in the 

 plagioclase, must have a Uttle lower crystallization-interval than 



^Jour. Wash. Acad. Set., Ill (1913). See also Day, Sosman, and Hostetter 

 {op. cit.), Amer. Journal. Set., XXXVII (1914). 



