of Minas Oeraes, Brazil. 23 



secondary (?) mica and, aside from the earthy oxides, a few rare 

 grains of tolerably large and well-formed zircons. This clay 

 body apparently represents an unsheared eruptive rock, possi- 

 bly of gabbroitic type. 



Still another clay horse some 2 meters thick is gneissic in 

 structure with white kaolinitic and greenish and yellowish 

 micaceous (?) elements and nodular masses of clay charged with 

 manganese and iron oxides. The residue after sliming consists 

 of secondary (?) mica without quartz, and heavy dirty white 

 earthy grains with occasional inclusions of rutile that are 

 almost certainly alteration products of ilmenite. Associated 

 with this layer is a thin one, a few centimeters thick, of a pul- 

 verulent clay heavily charged with graphite. 



With the exception of the above mentioned quartzose layer 

 (or sahlband), which apparently must be considered as an integ- 

 ral part of the ore body and which by its quartz and garnet 

 contents establishes a relation with the original type of the 

 Piquiry ore mass, the mineral at Sao Groncalo is so completely 

 altered by secondary processes as to give no clue regarding its 

 original character and origin. The included horses of clayey 

 matter may, a priori, be either inclusions of country rock, 

 segregated masses of the original rock of the ore body itself, 

 or intrusive dikes. The hypothesis may be ventured that the 

 gneissoid layer without manganese and with traces of original 

 ilmenite is of the first character, that the gneissoid layer above the 

 quartzose one and with traces of manganese oxide is of the second, 

 while the granitoid body with traces of manganese oxide and 

 with zircon may be a mass of the same character that has 

 escaped shearing, though it is more probably an intrusive dike. 



The ore of the main opening is almost exclusively a hard 

 stony, spongy psilomelane like that of Piquiry but presenting 

 more distinct evidences of shear structure. A prospecting 

 analysis by Mr. Scott gave metallic manganese 49*10 per cent, 

 siliceous residue 6*34: per cent, and phosphorus 0'126 per cent. 

 Lower down on the hillside another opening has been made 

 which affords in part a hard secondary ore of the same charac- 

 ter as the above, in part an ore of earthy aspect that was at first 

 considered to be of doubtful character but which on analysis 

 proves to be good merchantable stuff. This evidently repre- 

 sents an altered schistose (sheared) manganese-bearing rock in 

 which the resulting oxide has not, to any considerable extent, 

 migrated or been recrystallized but has apparently become 

 somewhat hydrated, since a rough test gives about 6 per cent 

 of water. Its appearance is that of a decomposed argillaceous 

 or calcareous schist profusely pitted with minute rounded cavi- 

 ties that are frequently lined with a fine white crust of second- 

 ary silica that, on dissolving the oxide, present the appearance 



