SEPTEMBER 17, 1897.] 
important contribution. Of much interest also 
are the chemical and genetic relations of this 
species as here presented; the crystals often 
enclose grains of what Professor Lewis terms a 
titanic spinellid, perhaps a titaniferous magne- 
tite, perhaps ilmenite, indicating a derivation 
therefrom’; they also, in some cases, lie in a 
curious manner, upon or around partially al- 
tered olivine crystals. The remark is made 
that, while perovskite is familiar in various 
non-feldspathie igneous rocks, it has not been 
found in peridotite until Professor Williams 
recognized it in the peculiar rock from Syra- 
euse, N. Y., and that later Professor Lewis 
identified it in the similar rock from Isom’s 
Creek, Kentucky, where it had been previously 
regarded as anatase. These three rocks, those 
just named and the African, are the only;known 
occurrences of what is here named Kimberlite. 
The article goes on to show that in basic 
eruptive rocks the titanium takes the form 
of perovskite, while in acid rocks it forms 
sphene; in intermediate ones it develops 
ilmenite or titanic iron; and these deductions 
harmonize precisely with important experi- 
ments of Bourgeois, in the artificial produc- 
tion of titanium minerals. 
After going into some particulars as to the 
minor minerals found in this rock, Professor 
Lewis then takes up the base or ground mass 
and discusses it minutely. He terms it ‘a 
more or less homogeneous serpentinous mass,’ 
very difficult to study by reason of its decom- 
posed condition, consisting now of a mixture of 
serpentine with calcite and some other prod- 
ucts of alteration, the original structure being 
wholly lost. 
Fragmental enclosures are frequent, ‘both 
of the adjoining shale and diabase, and also of 
more deeply seated granite, gneiss, eclogite, 
and other related rocks.’ Of these the shale 
predominates, sometimes making the rock a 
breccia. The shale itself is highly charged with 
carbon, so as to be quite combustible ; but the 
included fragments are altered, having lost their 
carbon and become harder, sometimes even 
metamorphosed to a micaceous structure, as 
before referred to. In size they vary from 
large masses, in the upper part of the mines, 
called by the workers ‘floating-reef,’ to small 
SCIENCE. 
453 
fragments, diminishing in number and size in 
descending. 
Professor Lewis goes into very detailed petro- 
graphical and chemical discussion as to the 
original character of the rock, in which it is 
hardly possible to follow him in a review, and 
finding no known rock that presents identical 
characters, he proposes for it the name of 
Kimberlite. This he designates as ‘a porphy- 
ritic volcanic peridotite of basaltic structure,’ 
and notes three forms of its occurrence: (1) 
Kimberlite proper, a typical porphyritic lava ; 
(2) Kimberlite breccia, the same rock broken 
and crushed by volcanic movements and 
crowded with included fragments of shale ; (38) 
Kimberlite tuff, the fragmental and tufaceous 
portion of the same rock. These varieties 
graduate into each other, and all occur together 
in the same neck or crater, the second, how- 
ever, being most abundant and most productive 
of diamonds. 
He treats of the origin of the brecciated 
structure, which has caused much discussion, 
some geologists regarding the whole rock as a 
sort of tufa or voleanic mud, while others hold 
that it isa true outpouring lava that has car- 
ried up fragments of the rocks broken through 
it its course, and has since been largely decom- 
posed. Professor Lewis urges the latter theory 
strongly, and supports it by many arguments ; 
while the editor, Professor Bonney, evidently 
inclines to the other view, advocated by Pro- 
fessor W. H. Hudleston, in 1885, and by 
some others. There is not space here to review 
Professor Lewis’ several arguments for the true 
igneous character of the Kimberlite and against 
the tufaceous theory. The one to which Pro- 
fessor Bonney accords the chief importance is 
the identity of the rock with that from Syra- 
cuse, New York, and Elliott county, Kentucky, 
where it occurs in actual dikes, such as are not 
found in tufas. The brecciated character, 
which is so marked, is referred by Professor 
Lewis to three causes, acting either separately 
or perhaps together. These are (1) rapid cool- 
ing and shrinkage; (2) ‘friction brecciation,’ 
from contact with the wall-rock ; and (3) sub- 
sequent movements and explosions in the crater 
itself, below. He illustrates and parallels the 
first of these from meteorites, to some of which 
