MAGMATIC DIFFERENTIATION OF IGNEOUS ROCKS 519 
per cent. With more basic plagioclase (andesine) somewhat 
more of the ferromagnesian silicate must be present if the latter 
is to commence crystallizing earlier than the plagioclase. 
2. In the granitic eutectic we only find a trifle MgO, viz., 
according to the analyses collocated on page 348, we find in magmas 
with little An only about o.2 per cent MgO (and in the magmas, 
somewhat richer in An, possibly as much as about 0.5 per cent 
MgO). 
The very great difference between the degree of solubility of 
ferromagnesian silicates (mica, pyroxene, or hornblende) in acid 
rocks, consisting chiefly of acid feldspar and quartz, and in basic 
rocks, containing basic plagioclase, is of great petrologic importance. 
ORTHORHOMBIC AND MONOCLINIC PYROXENE 
In orthorhombic pyroxene we often observe microscopically 
small lamellae of monoclinic pyroxene, and in monoclinic pyroxene 
corresponding minute lamellae of orthorhombic. This must 
(see above, p. 436) be explained as a secondary phenomenon, due 
to a secretion in the solid phase. 
Where independent individuals of orthorhombic and monoclinic 
pyroxene are intergrown, the orthorhombic (hypersthene)—as has 
often been pointed out by earlier investigators and as I have often 
observed in norites containing diallage—forms the kernel and 
the monoclinic (diallage) the surrounding parts. This indicates 
that the hypersthene was formed earlier than the diallage. Excep- 
tionally also we find large crystals of hypersthene, with quite good 
idiomorphic outlines, inclosed in the diallage. 
We here refer to Figures 20-21 of a norite,' containing diallage, 
from Skjekerdalen in Norway, and consisting of about 5 per cent 
olivine (according to the determination of the optical character 
and the axial angle with 20-25 per cent Fe,SiO,) in scattered 
individuals (not represented in the section drawn); about 25 per 
cent hypersthene (according to optical determination with about 
25 per cent FeSiO,); about 25 per cent diallage (optically positive, 
2V =ca. 65°, c:c=ca. 43°) with a little primary brown hornblende; 
* This rock has recently been treated by C. W. Carstens, Geology of the Trondhjem 
District, 1920, p. 101. 
