I go The American Geologist r March, i89t» 
from an olivinitic lava, and is not of the character of mud 
lava or tuff. Peridotytes resembling kimberlyte have been 
found in the United States at Syracuse, N. Y.,* and Elliott 
county, Kentucky .t So far as known no diamonds, however, 
have been found in the vicinity of these pcridotyte masses. 
None of the serpentines of California, so far as my informa- 
tion goes, are of the kimberlyte type. The association of the 
diamonds with serpentine may perhaps be due to the usually 
higher content of iron in these basic rocks. Diamonds might 
therefore be expected in any of the more basic igneous rocks. 
However, it is by no means certain that diamonds are not 
formed under very different conditions from those described 
by Crookes. In a recent paper* Derby writes of the occur- 
rence of diamonds in Brazil. Derby states that for the ques- 
tion of genesis the most significant of the Brazilian localities is 
that of Sao Joao de Chapada near Diamantina. The dia- 
monds occur here in thoroughly decomposed material, no ab- 
solutely fresh rock being found. So far as can be made out 
from observations on material the most unsatisfactory that can 
l)e imagined, the most plausible hypothesis as regards the va- 
rious clays of Sao Joao da Chapada is that they represent a 
group of phyllytes of varied character but principally if not 
exclusively of clastic origin threaded with dikes of pegmatyte. 
The clastic origin of the schists is regarded probable from the 
worn character of the zircons found in the clay derived from 
the schists. The diamond-bearing streaks appeared to Derb\' 
to contain distinct bands composed in part of quartz, with 
plates of specular iron, and these bands he suggests may be 
pegmat}'te dikes. The primar}' tourmaline and zircon and the 
secondary hematite and rutile found in the heav}' residue after 
washing the clay are supposed to have originated in the peg- 
matyte, the hematite, anatase, and rutile having formed from 
original iron and titanium minerals now gone. The staurolitc 
and disthene found in the heav}' residue are supposed to have 
come from the schists where they may have formed as a result 
*G. H. Williams, Am. Jour. Sci. Vol. XXXIV, 1887, p. 137. 
tJ. S. Diller, Bull. 38, U. S. Geol. Survey, 1887. 
JBrazilian evidence on the genesis of the diamond. Jour, of Geol. 
Vol. VI, 1898, pp. 121-146. 
