GEOLOGY OF CENTRAL MINAS GERAES, BRAZIL 399 



hydroxide pretty generally over the region and in some places the 

 building up of a series of beds of ferric hydroxide in a high state of 

 purity which sometimes totals several hundred meters in thickness. 

 The completeness of the separation, as evidenced by the fact that 

 much of the ore now contains 69 per cent iron out of a possible 70 

 per cent for chemically pure ferric oxide, and the vast extent of the 

 iron formation are very hard to understand. The iron oxide 

 resulting from the decomposition of granite, gneiss, amphibolite, 

 and crystalline schists should be in a more or less finely divided con- 

 dition, and as such should not be very generally separated from the 

 sand and clay by any sorting action of running water or ocean 

 waves, except in so far as affected by a higher specific gravity 

 which might result in small local accumulations of iron sands, or in 

 the association of smaller particles of iron oxide with larger particles 

 of other materials in the sediments. The iron oxide should be with 

 the schists and quartzites. Great masses of iron oxide containing 

 less than \ oi 1 per cent of impurity do not seem possible as the 

 result of mechanical separation, though it is not impossible that a 

 small percentage of the iron oxide in the iron formation may have 

 been derived thus. 



The alternative hypothesis is that the iron oxide was precipitated 

 from solution following the formation of limestones as an analogy. 

 This escapes the serious difficulties of separation of materials. As 

 in the case of the limestones, to account for the purity it is only 

 necessary to suppose that the deposition of the iron compound 

 occurred in clear seas where comparatively little clastic sediment 

 was being brought in. While the thickness of iron formation is 

 great, it is exceeded by many limestones. 



It is the nature of the precipitation that leaves the widest field 

 for speculation. There are two possibilities: (i) the iron may 

 have been precipitated directly from solution by purely chemical 

 means, or (2) it may have been abstracted from the water by organ- 

 isms and deposited as a sediment. In either case the iron may be 

 supposed to have been in solution chiefly in the form of ferrous 

 carbonate and to have been thrown down as ferric hydroxide. 



There are various chemical reactions which can cause the pre- 

 cipitation of ferric hydroxide from solutions of ferrous carbonate, 



