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



of the earth than in the nodules, while anatase only occurs spo- 

 radically. Zircon in extremely minute and usually ill-formed 

 grains of difficult identification was found in the micaceous 

 crusts but only sporadically in washings from other parts. 

 The phospbatic mineral is usually in ill-formed aggregates 

 that sink in the Klein solution, but a few rare crystals of 

 rhombohedral form were detected. By microchemical and 

 blowpipe tests (with the Florence bead) cerium was identified 

 in these grains and this with the rhombohedral form and high 

 specific gravity indicates that the mineral is the cerium-alumi- 

 num phosphate, florencite, recently described by Hussak and 

 Prior. Topaz could only be found in some but not all of the 

 nodular masses, but no trace of it could be detected in the main 

 body of the earth free from inclusions. When found it was 

 always in fragments of what had evidently been macroscopic 

 crystals, no perfectly formed microscopic crystals being 

 observed although they were carefully and confidently looked 

 for. 



The above observations afford a very insufficient basis for 

 the determination of the original type of the topaz-bearing 

 material, but in view of, the interest of the subject and the 

 slight probability of finding better preserved material for 

 study, some hypothetical deductions may be hazarded. In this 

 attempt at reconstructive geology two phases have to be con- 

 sidered, viz. the schistose one from which the earthy matter is 

 directly derived by decomposition, and the original rock type 

 from which this schist was derived by metamorphism accom- 

 panied by shearing. 



In the schistose phase it is certain that the rock was essen- 

 tially a micaceous schist, which, as the mica flakes, although 

 minute, are for the most part well formed, can be designated 

 as muscovite rather than as sericite schist. The mica contents 

 of this schist must have varied from about '70 per cent in the 

 normal parts free from segregations to about 30 per cent in 

 the more basic and to almost nothing in the more acid of the 

 latter. The iron and manganese contents varying in the 

 inverse proportion are probably referable to original oxides* 

 and in this case the schist must have been very similar in com- 



* Probably but not certainly, as in the immediate vicinity there are extensive 

 deposits of manganese ores that are residues from original carbonates and others 

 that are presumed to be derived from original silicates, while examples of the 

 complete replacement of an iron-bearing silicate (asbestiform amphibole) by 

 limonite are frequent in the neighborhood. In the case in question, however, 

 there are serious difficulties in the way of the hypothesis of the derivation of 

 these hydrous oxides from either a carbonate or a silicate. The other conditions 

 of the bed seem incompatible with the first, and the absence, at least apparent, of 

 secondary silica in the vicinity gives a strong though not absolutely conclusive 

 argument against the second. 



