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



and made analyses (reproduced in full in JN"o. 57 of this 

 Journal) proving that a great part at least of the uuctuous 

 schists of the region called talcose or chloritic by Eschwege 

 and others are relatively poor in magnesia and rich in alkalies 

 and are therefore essentially micaceous. Three of the schists 

 analyzed were from the topaz mine of Boa Vista or its imme- 

 diate vicinity, but it is almost certain that they represent what 

 Eschwege called argillaceous schist (fibrous schist of Gorceix) 

 rather than the topaz-bearing layers themselves, which are 

 nowhere found in a sufficiently sound state to give satisfactory 

 analyses. The lithomarge accompanying the topaz was also 

 analyzed, giving : Si0 2 ift'6 per cent, A1 2 3 38 per cent, MgO 

 1 per cent, and loss on ignition 14' 1 per cent — which is the 

 composition of a true kaolin with a slight percentage of mag- 

 nesia. The lithomarge is spoken of in one place as occurring 

 in a vein occupying a line of fracture but in another as result- 

 ing from an alteration of the schist. The alignment of the 

 principal mines noted by Eschwege was confirmed and some- 

 what extended by Gorceix and the existence of a second 

 parallel line marked by the smaller workings of Caxambu and 

 Fundao was affirmed. 



In recent brief visits to the Ouro Preto region an attempt 

 was made to resolve the doubts suggested by the above studies 

 and to determine the original character and mode of origin of 

 the topaz-bearing material. For this purpose the small mine 

 of Caxambu was selected, as here the topaz bed was tolerably 

 well exposed by recent workings, whereas in the larger mines 

 it is now for the most part concealed and everywhere greatly 

 obscured by landslides. A preliminary examination of these 

 mines fully confirmed the opinion of Gorceix of their sub- 

 stantial identity with that of Caxambu, which may therefore be 

 taken as typical of the topaz mines of the region. On account 

 of the extreme decomposition of most of the material to be 

 examined the problem presented was one of u mud geology" 

 — the attempt to reconstruct from earthy materials the original 

 rock types from which they were derived — and the solution 

 here presented is necessarily largely hypothetical. 



The mine is situated on the slope of a low col between the 

 base of the high-rounded campo-covered knob called the Morro de 

 Caxambii, and a lower knob a few score of meters to the south- 

 ward. In both of these knobs on each side of the col sound 

 ,rock is exposed in low bluffs. This is a sericitic phyllite 

 heavily charged with fine hematite dust and on the southern 

 side with a considerable amount of rather coarse quartz in scat- 

 tered grains, which gives the rock the appearance of an iron- 

 bearing quartzite (itabirite), though on examination it is seen 

 to be more micaceous than quartzose. The rock from both of 



