Chap. XHI.] 
BRAZILIAN DEPOSITS. 
273 
An interesting feature of the manganese-ores of the Vizagapatam 
district is the baryta found in the 22 analyses 
Baryta m t le ores. j-gferred to above. The amount of this constituent 
ranges from 0" 03 to 9 '53 per cent., the mean value being 2*03 per cent. 
Now according to the analysis of the sample of opalized kodurite, 
No. 233, given on page 257, and carried out by Messrs. J. and H. S. 
Pattinson, kodurite is to be supposed to contain either no barium at 
all, or such a small quantity that it cannot be estimated. Mr. J. Coggin 
Brown, however, found 0* 18 per cent, of BaO in the specimen of spandite 
that he analysed, and possibly all spandite if carefully tested would be 
found to contain a trace or small quantity of this constituent. Since the 
amount of BaO in the original rocks is so small, the comparatively 
large quantity found in the ores cannot be simply residual, i.e., have 
been left behind with the oxides of manganese and iron and a certain 
proportion of the AI2O3, when the garnets suffered decomposition. 
It is necessary to suppose that this was one of the constituents that 
went into solution, probably as bicarbonate ; it must have been thus 
concentrated and later deposited during the formation of the manga- 
nese-ores. Unless, however, a much larger proportion of manganese 
passed into solution than was ultimately deposited as manganese-ore, 
it is diflficult to explain the great increase of the BaO relative to the 
manganese oxides. If the supposition put forward in the former part 
of the previous sentence be thought to be unhkely, then it is necessary 
to suppose that the waters that originally attacked the masses of 
kodurite rocks brought in the barium oxide as a part of their burden, 
having obtained it from a source different to the rocks of the kodurite 
series. 
It will be interesting to compare the^^ Indian manganese-ore deposits 
Comparison with the oi the kodurite series with those of the Queluz 
Brazilian deposits. district. State of Minas Geraes, Brazil, the only 
ones, as far as I can discover, that bear any resemblance to 
those of Vizagapatam. Dr. Orville A. Derbyi, who has studied the 
Brazilian deposits, ascribes their formation in many cases to the decom- 
position and leaching of basic manganiferous rocks of which the most 
important mineral is manganese-garnet (spessartite). These masses of 
manganiferous rock often occur as dyke-hke masses and are considered 
to be the result of magmatic segregation from a basic magma of dioritic, 
gabbroitic or noritic tj^e, now represented by the very decomposed, 
Amer. Jour. Set. XII, pp. 18-32, (1901). 
