Derby — Accessory Elements of Itacolumite, etc. 187 



Art. XXVI. — On the Accessory Elements of Itacolumite, and 

 the Secondary Enlargement of Tourmaline ; by Orville 

 A. Derby. 



The peculiar type of granular quartz-rock characteristic of 

 the gold and diamond regions of the Serra do Espinhaco in 

 the state of Minas Geraes, Brazil, which has come to be gen- 

 erally known by the name of itacolumite, proposed in 1822 

 by Eschwege,* has given rise to much discussion. Eschwege 

 and the majority of subsequent writers on the geology of the 

 region (Helmreichen, Claussen, Pissis, Heusser and Claraz) 

 regarded it as a member of the primitive group of rocks 

 having a special mode of formation, an idea that has found 

 expression in some recent text-books of geology (Lapparent, 

 Traite de Geologie, 2d ed., Paris, 1885). Later writers (Hartt, 

 Liais, Gorceix, Derby) have regarded it as a metamorphosed 

 sandstone. The view that it is only a special phase of an ordi- 

 nary sandstone has also been presented, though not, so far as 

 known, by any geologist who has seen it in place. The most 

 definite evidence thus far presented by the advocates of any 

 one of these discordant views is that of the very pronounced 

 and characteristic schistose structure by those who maintain 

 the metamorphic hypothesis, but the force of this argument, 

 which was long considered as conclusive in favor of a clastic 

 origin, has been greatly weakened by recent studies on the 

 crystalline schists. 



On the hypothesis that, if clastic as presumed, some internal 

 evidence of the mode of origin might be found in the mineral 

 elements of the rocks themselves, a microscopic examination 

 has recently been made of the isolated grains separated by 

 washing from hand specimens of the rock, from several of 

 the typical localities in Minas Geraes as well as from 

 other points in the same state and from those of Goyaz and 

 Bahia. As the original itacolumite of Eschwege was divided 

 by him into a schistose and massive group, which, by later 

 writers (Derby, Gorceix), have been considered to be inde- 

 pendent formations, it may be stated that the material exam- 

 ined was for the most part known, or presumed, to represent 

 the schistose and older group, and that in several instances the 

 specimens presented the property of flexibility which was for 

 a long time, but erroneously, supposed to be characteristic of 

 the rock as a whole. 



The quartz grains of these rocks present no unequivocal 



* Geognostiches Gemalde von Brasilien, und wahrscheiuliches Muttergestein 

 der Diaraanten. Weimar, 1822. 



