Diamond Regions of Eastern Minas Geraes, Brazil. 213 



the peculiar manner in which the rock has been invaded by 

 tourmaline. This mineral besides lining the shear and fracture 

 planes with beautiful and symmetrical microscopic rosettes of 

 radiating needles, also appears in the body of the rock, where, 

 however, it is confined to the white areas, the needles being 

 sharply cut off by the rectangular limits of the original crystals 

 wholly or in part substituted by tourmaline. This invasion 

 evidently took place after the decomposition and shearing but 

 perhaps daring the process of the recrystallization of the rock. 



There is nothing intrinsically improbable in the hypothesis 

 that some metamorphic schists may result from the alteration 

 of rocks, both massive and clastic, that had been decomposed and 

 leached in situ, and thus many of the anomalies of composition 

 that this group presents may be satisfactorily explained. This 

 hypothesis is involved in that presented by Vogt and Yan 

 Hise of the derivation of micaceous iron schists of Norway 

 and of the Lake Superior region from original carbonates, and 

 in another place evidence pointing in the same direction will 

 be presented in relation to the extensive itabirite beds asso- 

 ciated with the schists here discussed. If, as seems probable, 

 the itabirites are derived from decomposed and leached car- 

 bonates, the process must have taken place on a gigantic scale 

 in the region in question and the associated silicate rocks, 

 whether massive or clastic, must have been affected to a greater 

 or less extent. 



The hypothesis of leaching cannot, however, be applied sat- 

 isfactorily to the rock No. I, which, if an original basic 

 eruptive of any of the ordinary types, must have lost heavily 

 in lime and iron while magnesia was retained. This, however, 

 would be contrary to what is supposed to have taken place in 

 the other rocks here discussed and to what is generally observed 

 in the modern cases of decomposition and leaching. Moreover 

 the presence of monazite in this rock is, so far as present expe- 

 rience goes, incompatible with such types. It seems more 

 probable that in this case the composition has remained com- 

 paratively unchanged and that the original rock must have 

 been one of low silica, iron and lime, but of high alumina and 

 magnesia associated with a considerable variety of rare ele- 

 ments. No type of rock with these characteristics has as yet 

 been clearly defined, but that such may exist, either indepen- 

 dently or as local phases or segregations in other types, is sug- 

 gested by the occurrence of corundum and aluminous silicates 

 in association with olivine rocks in North Carolina and Geor- 

 gia, and that of alumina-magnesia silicates such as cordierite, 

 prismatine and sapphirine as segregation (?) masses in the gran- 

 ulite of Saxony and the mica schist of Fiskernas, Greenland. 

 If, as may be presumed from the scanty information at hand, 



