32 Derby — Occurrence of Topaz near Ouro Preto, Brazil. 



ing of metallic oxides in certain nodules and about certain 

 others becomes plausible. All things considered, however, it 

 seems more probable that this last is rather due to an original 

 and not a secondary feature. 



The question of the original type from which, by shearing 

 and metamorphism, this muscovite-iron-manganese schist was 

 produced, is a more difficult one. Whether the blue phyllite 

 was originally eruptive or not, there can be little doubt that 

 the topaz-bearing schist enclosed in it was so. This is indi- 

 cated by the character of its contacts, by the absence of clastic 

 elements, by its manganese contents, and by its banded and 

 nodular structure. Being eruptive, it must have been a rock 

 without free quartz but with a comparative abundance of 

 alumina-alkali silicates (feldspars or feldspathoids) and of iron 

 and manganese minerals, which last, if in the state of silicates, 

 must have been bisilicates containing lime and magnesia that 

 have disappeared by leaching either in the recent decomposi- 

 tion, or, as seems more probable, before the metamorphism 

 and shearing. The rock was probably porphyritic in structure 

 (perhaps not necessarily so if an ancient period of decomposi- 

 tion be admitted) and full of segregated masses of both the more 

 acid and the more basic constituents, in the former of which 

 quartz, not a normal constituent of the rock, also appeared. 

 These segregations must have been more or less drusy and in 

 them several minerals, of which some were not normal to the 

 rock, crystallized with one free termination. 



The conditions above enumerated point to some member of 

 the augite- or nepheline-syenite groups as the most probable 

 original eruptive type. The rocks of these groups would 

 afford the necessary alumina and alkalies to furnish the abund- 

 ant mica free from quartz that characterizes the metamorphic 

 schist, while their more or less basic phases would afford the 

 metallic oxides as well, if the accompanying lime and magnesia 

 can be supposed to have disappeared by leaching. Moreover 

 these rocks are particularly subject to rapid alternations in 

 texture or composition, or both, giving segregated masses, or 

 Schlieren. In those known to me in various Brazilian locali- 

 ties both acid and basic segregations occur, and frequently one 

 of the former character is surrounded by a zone more basic 

 than the normal rock, thus giving the conditions noted in some 

 of the topaz nests. 



In this connection it is interesting to note that the study of 

 the more normal schists of this and of the Diamantina region 

 led to a similar conclusion, that rocks of the augite- or nepheline- 

 syenite groups were probably represented among them, though 

 in this case it seems necessary to admit that clastic as well as 

 eruptive types occur. Combining the two series of observa- 

 tions, the hypothesis may be ventured of the existence in this 



