of Minos Geraes, Brazil. 21 



more heavily charged with graphite. The very abundant resi- 

 due obtained by washing, or by treatment with acid, is in great 

 part of a dirty white color, the garnet grains being discolored 

 by a closely adherent graphitic powder. The amphibolic min- 

 eral, apparently more abundant than in ~No. I, appears in the 

 residues in rudely prismatic forms but is too heavily stained 

 with manganese oxide to permit of a satisfactory determination. 

 No apatite could be detected in the residue, but the acid solu- 

 tion from the rock gave a distinct reaction for phosphoric 

 acid. 



3d. A milky white rock which under the microscope is seen 

 to be composed of about equal parts of garnet and quartz, the 

 latter being abundantly threaded with delicate transparent 

 needles of a white asbestiform mineral. The garnet is for the 

 most part larger and better crystallized than in the specimen 

 above described and has a decided yellow tinge. The quartz 

 in a tine mosaic about the garnet grains and in minute refilled 

 joints is almost certainly secondary, filling the place of some 

 mineral that has disappeared. The asbestus (?) needles, in part 

 free, in part included in the quartz, are undoubtedly secondary 

 but not, as in the case above described, formed without migra- 

 tion by the transformation in place of some preexisting min- 

 eral. The only accessory, embedded in the garnet as well as in 

 the quartz and thus probably of primary origin, is a transpar- 

 ent red mineral in minute grains and hexagonal flakes that 

 give strong reactions for both titanium and manganese. The 

 crystallographical, optical and chemical characters of this 

 mineral, so far as they can be made out, agree with those of 

 pyrophanite. 



The above observations indicate that the original rock from 

 which this ore body was derived was essentially a manganese 

 garnet rock containing sporadically (and perhaps in segregated 

 masses) an amphibole mineral, apatite, a titanium mineral and 

 presumably an easily decomposable silicate that has entirely 

 disappeared. Graphite is also distributed capriciously through- 

 out the mass, but, as will be shown below, this is perhaps not 

 an original or essential element. The predominant quantity of 

 merchantable ore of high grade shows that by far the greater 

 part of the mass must have been an almost absolutely pure 

 manganese garnet rock from which silica and alumina have 

 been leached out. Since in the process of oxidation iron oxide 

 would almost certainly have remained and have been concentrat- 

 ed with that of manganese, the original rock (and especially the 

 garnet) must have been notably free from this element, which 

 in the ore analyses is in smaller proportions than in the rock 

 sample above analyzed. It is worthy of note that no free iron 



