THE MICAS, ASBESTOS, AND GEMS 163

have mostly been obtained where they have been reconcentrated
 in river gravel; they are associated with quartz,
zircon, tourmaline, disthene, and mica; the only igneous
rocks known are the pegmatites in the underlying schists
and occasional pebbles in the conglomerate, and a kimberlite
tuff at Uberaba, which contains no diamonds (Hussak,
Z. prakt. G., xiv, 1906, pp. 322-4).
The younger Brazilian deposits, the Levras sandstone and
conglomerate, of Bahia, may be Carboniferous in age. The
associated minerals include tourmaline, zircon, garnet,
staurolite, and kyanite. - No igneous rocks occur with the
diamonds, which have been attributed to undiscovered
dykes by D. Draper (Mining Mag., ix, 1913, p. 435), E. C.
Harder and R. T. Chamberlin (Yourn. Geol., xxiii, 1915,
P. 418), and Miller and Singewald (Min. Dep. S. Amer.,
1919, pp. 213-4). On the other hand, J. C. Branner (Amer.
Fourn. Sci., (4), xxxi, 1911, p. 490) regarded the diamonds as
formed during the metamorphism of the quartzites, and
according to O. A. Derby (Journ. Geol., xx, 1912, p. 455) they
were due to pneumatolytic alteration of a fractured ultrabasic
 rock. The mineral association suggests a pneumatolvtic
origin by some ascending boric acid solution.
The diamonds discovered in 1908 along the coast of Southwestern
 Africa were at first regarded as washed out of submarine
 necks of kimberlite. According to E. Kaiser (Diamantenwiiste
 Sud-West- Africas, 1926, ii, pp. 329, 338) they came
from various sources, and were collected into a sheet of sandstone,
 whence they were carried into Eocene and later sands,
The association of these diamonds with native gold and copper,
chalcopyrite and pyrite, iron and manganese oxides, garnets,
zircon, sillimanite, tourmaline and topaz, does not suggest
an ultra-basic igneous origin ; some were probably formed in
contact zones with the sillimanite, and others under pneumatolytic
 conditions with tourmaline and topaz. The diamonds
 are small ; the largest found weighs only 50 carats.
The diamonds of the field on the Gold Coast discovered
by Sir A. E. Kitson (Gold Coast G.S., Bull. i, 1925, p. 35) are
attributed to the action of igneous rocks on carbonaceous
slates.
TuEORIES OF FORMATION BASED ON SUPPOSED ARTIFICIAL
Diamonps—The current theories as to the formation of