Crosby.] 486 {April 7, 
white, coarsely crystalline, and contain a small amount of green ser- 
pentine. Specimens of the serpentinic limestone have been sub- 
mitted by Mr. Derby to Dr. J. W. Dawson,! and this high authority 
announces that they contain Eozoon and are of Laurentian age. 
Although the reference of this older series of Brazil to the Lauren- 
tion is undoubtedly correct in the main, yet it is likely that future 
investigation will require farther subdivision. For in the descriptions 
of Darwin and some other writers, and especially of Liais, there are 
many passages referring to micaceous gneisses and schists containing 
garnets and allied minerals, and otherwise indicating the exist- 
ence among the older crystallines of Brazil and Uruguay of typical 
Montalban strata. In fact, any one familiar with the mineralogical 
characteristics of the Laurentian and Montalban systems will recog- 
nize that Mr. Derby himself has set forth this division very clearly in 
his last contribution to the subject,? where he says, ‘‘ The succession of 
the Brazilian gneisses is, according to the observations of Pissis, 
Hartt, Liais, and myself, (1) porphyritic or granitic gneiss, with 
red feldspar ; (2) fine-grained gray gneiss (leptinite of Pissis and 
Liais), often garnetiferous and schistose; (3) fine-grained schistose 
gneiss, passing to mica schist, with subordinate beds of quartzite, 
and an abundance of mineral veins not found in the lower parts of 
the series. This last, the upper or metalliferous series of Pissis and 
Liais, prevails in the Mantiqueira Range, while the lower divisions 
predominate in the Serra do Mar and Parahyba Valley.” 
The line between divisions (1) and (2) corresponds to a marked 
contrast in mineral characters, and appears to separate the Lauren- 
tian below from the Montalban above ; and the division, as thus drawn, 
appears to be as valid Si dheapienne as mineralogically. 
The extensive deposits of petrosilex (porphyry) and siramhed 
hornblendic rocks (greenstones) noted by Darwin and others in 
southern Brazil may, with much probability, be taken to represent 
the Huronian system. On the Pacific slope of the Andes, from Pat- 
agonia to Peru, or farther north, the older formations show a similar 
three-fold division, the characteristics of the Laurentian, Huronian 
and Montalban systems being, apparently, strongly marked and eas- 
ily traced. 
The second great division of the Eozoic rocks of Brazil is undoubt- 
edly newer than, and, in the opinion of Mr. Derby, reposes uncon- 
1 American Journal Science xIx, 1880, pp. 324-327. 2Thbid. 
~~ deg 
alec ami yi 
- ~~ Sat reatN 
DRM i Ho eo 
