THE MATRIX OF THE DIAMOND 35
Perovskite
has not been identified in a peridotite until quite recently, when my
friend, Dr. G. H. Williams, observed it in an eruptive
peridotite—closely resembling the Kim-berley rock—in the State of New
York. It has probably been mistaken frequently for other minerals. The
colour and form, the similarity to picotite, the included opaque
grains, and the mineral associations, all indicate perovskite rather
than zircon. Diller1 has described, as probably anatase,
certain yellowish, highly-refracting grains occurring around and
penetrating ilmenite in an eruptive porphyritic peridotite in Elliott
County, Kentucky. Through the courtesy of Mr. Diller and of Prof. A. E.
Crandall, I have received specimens of this interesting rock, and find
that these grains are identical with the perovskite of the Kimberley
rock.2 The same form, colour, and enclosures, the same
twinning structure, the same high index of refraction and low power of
double refraction, and the same mineral associations, occur in the
Kentucky as in the Kimberley peridotite.
It
is an interesting fact that, speaking generally, the titanium in acid
eruptive rocks takes the form of sphene, in basic non-felspathic rocks
of perovskite, and in rocks of intermediate basicity, the
felspar-basalts, of ilmenite or titanic iron. An explanation is offered
by an important experiment performed by Bourgeois,3 who was
able to form artificial perovskite by fusing its elements with various
silicates and basic rocks, and who found that below a certain point of
acidity only perovskite was formed, while above that point only sphene
was produced. With a titanic spinellid as the primary titanium mineral,
the reaction of a basic magma would produce perovskite (as in
leucite-and melilite-basalts and in peridotite), while the reaction of
an acid magma would produce sphene (as in granite,
