0. A. Derby — Magnetite Ore Districts in Brazil. 313 



planes, which give it the aspect of a mica schist. Sections 

 from the inner portion of one of these mica-covered slabs (the 

 outer micaceous portion is too friable for section cutting), show 

 the rock to be composed almost exclusively of irregular grains 

 of violet-colored titaniferous pyroxene with, as accessories, an 

 abundance of iron minerals (magnetite and titaniferous iron), 

 rare grains of perofskite and milky white grains, in part apatite, 

 in part a zeolitized silicate. No mica can be seen in these 

 sections. Higher up in the cuttings where the rock is so 

 friable as to readily crumble in the lingers, the mica becomes 

 more abundant, lining not only the original planes of schistosity 

 but all secondary divisional planes including those between the 

 partially disaggregated grains of pyroxene. The origin of the 

 mica from the weathering of the pyroxene is perfectly clear, 

 and higher up in the cuttings the rock is seen to lose entirely 

 its lithoid character, and to pass to a yellow clay composed ex- 

 clusively of decomposed mica flakes.* This in turn passes to 

 a dark-red soil which, like the yellow micaceous clay affords, 

 on washing, the characteristic residue of the sound pyroxene 

 rock, viz : iron minerals, apatite and perofskite. 



At many points throughout the district, the iron minerals in 

 the rock above described rise from the role of an accessory to 

 that of an essential element giving layers composed of various 

 proportions of pyroxene and magnetite which, by the dis- 

 appearance of the pyroxene, pass to a pure iron ore. This 

 enrichment in iron is accompanied by the total disappearance 

 of the white silicate, and a diminution in the quantity of perof- 

 skite. In specimens of nearly pure ore, the magnetite forms a 

 continuous network enclosing detached grains of pyroxene in 

 the same manner as the me- 

 tallic portion of a meteorite 

 of the mesosiderite type en- 

 closes the silicate portions. 

 The annexed figure from a 

 microphotograph of a sec- 

 tion of Jacupiranga ore rep- 

 resents this structure. In 

 one place a layer about 15 

 centimeters thick of pure 

 magnetite containing pyrox- 

 ene (in part altered to mica) 

 only in a thickness of a few 

 millimeters at the upper and 

 lower surfaces, was seen in 



* This wholesale passage of pyroxene to mica through weathering is a capital 

 feature in the geology of the two districts and has been very carefully verified. 

 It is not confined to any particular kind of pyroxene as it has been noted also in 

 acmite-bearing rocks and in the pyroxene rocks associated with marble beds. 



