MAGMATIC DIFFERENTIATION OF IGNEOUS ROCKS 635 
discontinuous mix-crystals with a eutectic boundary consisting of 
about 2.5-3 per cent spinel:97.5-97 per cent magnetite (cf. my 
treatise, cited p. 2, “Uber das Spinell: Magnetit-Eutektikum,” 
IQIO). 
In the igneous rocks, containing primary spinel, this mineral, 
as is well known, belongs to the very first stage of crystallization 
when it is present in at least about o.1 per cent by weight. But if 
the quantity of spinel sinks still lower the case is different. As an 
example we may mention that in the olivine-hyperite, illustrated in 
Figure 33, where first some olivine crystallized and later a little 
magnetite, some quite small individuals of spinel often appear in 
the magnetite. The spinel forms about 1 per cent of the magnetite, 
and since the latter forms 1-2 per cent of the entire rock, the 
quantity of spinel may consequently be set to o.o1-0.02 per cent. 
And this minimal quantity of spinel did not crystallize until 
mineral No. 1, in this case the olivine, had commenced forming. 
Further we shall mention an ilmenite-norite from Storgangen 
near Soggendal (Ekersund), consisting of about 40 per cent ilmen- 
ite, 40 per cent hypersthene, 20 per cent labradorite, and in addition 
I—2 per cent biotite, about o. 2 per cent pyrite, and about o. 1 per cent 
spinel. The latter is intensely green and must be termed hercynite. 
The pyrite appears in small cubes (with edges o.1-o.3 mm.) 
and on these quite small spinels (o.04-o.1 mm. large) have grown. 
These were deposited on the crystals of pyrite,t which in this 
manner served as /ixkorper in the still molten magma. The 
small spinels show an octahedron boundary against the surround- 
ing, later mineral—ilmenite, labradorite, or hypersthene. Fes, 
as well as (Mg, Fe) (AL, Fe.)O, were so little soluble in this magma © 
that, although present in so small a quantity, they commenced 
crystallizing at an earlier stage than the silicates and the iron ore, 
but in such a manner that at the very first stage some pyrite solidi- 
fied, and then, at somewhat lower temperature, the spinel. 
The essential part of the apatite crystallizes, as is well known, at 
a very early stage. Sometimes, in the same thin section (Fig. 35), 
we may find a crystal of pyrite serving as Fixkérper for the deposit 
t The same sequence of crystallization, spinel later than pyrite, I have mentioned 
also in an earlier treatise, see Zeitschr. f. prakt. Geol. (1900), p. 238, Fig. 39. 
2 
