84 Washington — Pyroxenite and Hornblendite in Brazil. 



and the presence of titanium, etc., the composition much 

 resembles that of the hornblende of the Maracas rock. 



Pyroxenite. 



This rock forms the outer portion of the mass. The transi- 

 tion between the two types is gradual, and apparently the 

 quantity of pyroxenite is much greater than that of the horn- 

 blende, though no definite estimates were made. The rock is 

 distinctly foliated parallel to the contact with the gneiss. 



The pyroxenite is a glistening dark brownish black, rather 

 coarsely granular, made up largely of bright glistening pris- 

 moids and grains of clove-brown pyroxene from 1 to 2 uim in 

 diameter. It is distinctly finer grained than the peridotite. 

 Grains of greenish black hornblende are also present, but they 

 are not conspicuous as compared with the pyroxene. The rock 

 is evidently very fresh, except near the contact with the gneiss, 

 where it has undergone some alteration and impregnation with 

 copper minerals. 



In thin section the rock is seen to be noncrystalline and 

 with granitic texture. The dominant pyroxene is orthorhom- 

 bic. It may be called hypersthene rather than bronzite, fol- 

 lowing Dana and Hintze, as it contains about 12>% of FeO and 

 is pleochroic. It forms thick, stout, rather prismoidal subhe- 

 dral as well as anhedral grains. It is of a very pale pinkish 

 brown or flesh color, and shows decided but not very strong 

 pleochroism; «, flesh pink, /3 and 7 colorless. Rather coarse 

 cleavage lines are prominent and the extinction is always paral- 

 lel to these. The pyroxene is perfectly clear and fresh, and 

 carries no inclusions. A very pale olive-green hornblende is 

 less abundant than the hypersthene. It forms usually anhedral 

 grains, or very rarely, prismoids. In its general character it 

 resembles that of the hornblendite, but the color is paler and its 

 pleochroism extremely weak. It also is perfectly fresh and 

 with fine cleavage lines well marked. The crystallization of 

 the hypersthene and hornblende would seem to have been in 

 great part contemporaneous, as each occurs more or less par- 

 tially enclosed in or molded against the other, as well as in 

 anhedral grains ; but on the whole much of the pyroxene seems 

 to antedate the hornblende, some of which is evidently the last 

 product of crystallization. 



There are a few small anhedra of colorless olivine which 

 show signs of slight incipient alteration in narrow yellow bor- 

 ders here and there. Small anhedral grains of a dark green 

 spinel, probably pleonaste, are rather common, but not a grain 

 of magnetite was seen. There is absolutely no feldspar. 



Chemical Composition. — An analysis made by me is given 

 below, along with two others of closely similar rocks. This 

 analysis resembles those of several other pyroxenites, two of its 



