Washington — Pyroxenite and Hornblendite in Brazil. 89 



experiments on the Bahia hornblendite. Unfortunately, the 

 time at our disposal was rather limited, owing to mj impend- 

 ing departure for Naples, and the results are not wholly in 

 harmony with those of Lacroix. Further work along these 

 lines will be undertaken later. 



On melting a rather large sample of the hornblendite in a 

 gas furnace, and cooling slowly, a noncrystalline mass was 

 obtained. This showed in thin section many very thin plates, 

 appearing as long thin needles in the section, of a colorless 

 monoclinic pyroxene with extinction angles up to about 30° 

 and with rather indefinite and fuzzy borders. There are also 

 many similar thin plates of a colorless olivine or possibly ortho- 

 rhombic pyroxene with sharp and clearcut borders and many 

 phenocrystic grains of magnetite. These lie in a greenish, 

 somewhat cloudy groundmass which is resolved by high powers 

 into a felt of very small prisms of greenish augite, and many 

 minute grains of magnetite. The augite prisms here and there 

 are arranged in branching or feather-like patches. The base 

 in which these are imbedded appears to be in great part holo- 

 crystalline, colorless and with lower refractive index than 

 pyroxene. It appears to be feldspar, in part at least, though 

 the absence of twinning lamellae and crystal forms, and the 

 difficulty of determining the optical constants, render this 

 identification somewhat uncertain. 



Another sample of the hornblendite, on heating in an elec- 

 tric furnace, was found by quenchings to have a melting inter- 

 val' of from about 1200° to 1400°. After heating to 1450° 

 (that is, above its melting point) for some time, and then being 

 cooled slowly to 1250° and held at this temperature for 5 

 hours, thin sections show beautifully sharp phenocrysts of a 

 colorless olivine. Dr. H. E. Merwin kindly examined these and 

 found the refractive index to be y = 1*680, a = 1*643, =h '003, 

 corresponding to forsterite. They are somewhat skeletal in 

 development, carrying elongated central cores of greenish 

 groundmass. They are elongated parallel to a pinacoid and 

 show sharp domal terminations at either end, the domal angle 

 varying between 80 and 90°. No phenocrysts of monoclinic 

 pyroxene were seen. The groundmass here is yellowish-green, 

 somewhat resembling the other. It is composed largely of 

 minute prisms of augite, not felted as in the previous case but 

 arranged in oblong featherlike aggregates, each small patch 

 outlined by a narrow row of minute magnetite grains. The 

 basis between these could not well be made out, but there 

 seems to be nothing which can be considered to be feldspar, 

 and the small aggregates appear to be wholly of an augitic 

 pyroxene. These results are, as has been said, somewhat 

 inconclusive, and the evidence is not strongly in favor of the 



