the Secondary Enlargement of Tourmaline. 189 



fair size and abundant, the signs of wear are generally so dis- 

 tinct as to be unmistakable. The predominant type is that of 

 short, stout, rounded prisms, frequently of a rose color and 

 sometimes partially whitened (by incipient malaconization in 

 the parent rock?), which is suggestive of granite or granitic 

 gneiss as the original rock from which they were derived. 



The iron ores, magnetite and hematite are usually present, 

 though often in very minute quantities. The two oxides some- 

 times occur together, though in general hematite is the most 

 abundant and characteristic. Both are sharp-angled with 

 lustrous faces and are evidently audiogenic. Pyrite is fre- 

 quently present and often in octahedral forms which on becom- 

 ing altered to limonite (or hematite?) afford martite-like forms 

 difficultly distinguishable from those derived from magnetite. 

 Ilmenite has not been observed, though titanium oxide in the 

 form of isolated crystals or aggregates of rutile or anatase, or 

 both, is frequent and always with the perfectly fresh appear- 

 ance of an audiogenic element. Considering these rocks, on 

 the evidence of the zircons, as original sandstones derived from 

 granites or gneisses and with a certain concentration of the 

 heavy accessories of the parent rock, the presumption is that 

 the iron ores would be equally or more abundant than the 

 zircons and be represented by magnetite, titaniferous magne- 

 tite or ilmenite. On this hypothesis, there has been in the 

 metamorphism of the rock a notable reduction in the quantity 

 of the free iron oxides, a recrystallization of the remaining 

 portion as hematite and more rarely as magnetite, and an isola- 

 tion of the titanium oxide when present as rutile or anatase. 

 The iron oxide that has disappeared as such has perhaps gone 

 to the green mica, or the pyrite, or possibly has been leached 

 out. In connection with the supposed isolation of the titanium 

 from ilmenite, an iron sand from the diamond region of Len- 

 coes in Bahia is interesting. Agglomerated nodules of worn, 

 or etched, grains of iron oxide giving the characteristic streak 

 of hematite show perfectly fresh microscopic rutile in the 

 interspaces between the grains. They are presumed to come 

 from the sandstone formation of the region and to represent 

 an ancient natural concentrate of black sand (magnetite, per 

 haps titaniferous, and ilmenite) in the original sediments. As 

 is well known, the transformation of magnetite to hematite and 

 the formation of rutile and anatase from titanium-bearing 

 minerals (sphene or ilmenite) may take place in the process of 

 atmospheric weathering, and are not therefore necessarily indic- 

 ative of metamorphism. When, however, as in the case of 

 the rocks in question, the hematite takes the form of a well- 

 crystallized specular iron, it may be presumed that true meta- 

 morphic changes are involved. 



