of Minas Geraes, Brazil. 29 



resistant and differs somewhat in aspect from that of the first 

 opening but like that gi ves only a residue of ilmenite and appears 

 to have been an amphibole schist. 



Some 50 to 60 kilometers to the southward of Queluz and in 

 the municipal district of Barbacena is another ore district 

 which I have not had an opportunity of visiting but from 

 which specimens from various points have come to hand. 

 These represent an extension of about 30 kilometers as meas- 

 ured by the railroad line between the stations of Ressaquinha 

 on the north and Sitio on the south. The region, like that 

 about Queluz, is characterized by gneissic rocks abundantly 

 injected with granite. The ores are of two types, of which 

 one, corresponding to that of the Queluz district, consists of a 

 garnetiferous rock impregnated with secondary manganese 

 oxide evidently derived from the garnet. One specimen is 

 heavily charged with well-crystallized graphite and gives in 

 the residue a white amphibole, neither of which minerals have 

 been noticed in the other specimens. The second type is a 

 manganiferous magnetite of which a specimen from near the 

 station of Pessaquina was analyzed by Mr. Scott, who found 

 11*60 per cent of metallic manganese with 40-08 per cent of 

 metallic iron. The other specimen at hand seems to be some- 

 what richer in manganese but is still essentially an iron-man- 

 ganese ore. The appearance of this ores is that of an ordinary 

 finely granular magnetite charged with pulverulent secondary 

 manganese oxide. The origin of this oxide is readily found by 

 dissolving the metallic oxide with hydrochloric acid, which 

 leaves a more or less abundant residue of rather coarse and 

 highly corroded spessartine often reduced by superficial etch- 

 ing to irregular hook-shaped fragments. The original type 

 was therefore a magnetite-spessartine rock from which the 

 silica and alumina of the garnet has been almost completely 

 removed by leaching, leaving a residue of manganese oxide. 

 From another ore district near Paranagua in the state of 

 Parana a specimen of identical appearance giving about 12 per 

 cent of metallic manganese is at hand, which appears to repre- 

 sent the same type from which the original garnet has entirely 

 disappeared. This conclusion is confirmed by a specimen 

 received later from the state of Santa Catherina, but in the pro- 

 longation of the Parana ore district, in which, as in the Barba- 

 cena ore, a remnant of corroded spessartine is still preserved. 

 In this costal ore district, extending from southern Sao Paulo 

 to Santa Catherina and embracing the Jacupiranga deposits 

 described by me some years ago in this Journal (April 1891), 

 titaniferous magnetites are very abundant and characteristic, but 

 their relations to these rarer manganese-magnetite ores are not 

 known. 



