GEOLOGY OF CENTRAL MINAS GERAES, BRAZIL 403 



tion, the field evidence appears to indicate that the deposition of 

 the iron-oxide beds in these locaHties of unusual thickness took 

 place with comparative rapidity. 



While certain thin discontinuous layers of iron formation 

 might be attributed to chemical percipitation, it is difficult to 

 realize how great thicknesses over large areas could have been thus 

 formed. Little is known concerning the concentration of iron 

 compounds in solution necessary for chemical precipitation. It 

 probably varies greatly with varying conditions, and the special 

 conditions which would have to be assumed as causing the 

 deposition of the Brazilian iron formation must have extended 

 over large, as well as widely scattered, areas since this formation is 

 not only found throughout a considerable area in Minas Geraes 

 but also exists in the extreme western part of Brazil, nearly a thou- 

 sand miles to the west. On the other hand, there is experimental 

 evidence that bacteria do precipitate iron oxide out of very dilute 

 solutions, and it is only necessary to assume the presence of large 

 numbers of these micro-organisms in scattered localities to account 

 for the presence of the iron formation. Unfortunately the meta- 

 morphism which the iron formation has suffered makes it impossible 

 to recognize organic remains, if such were originally present. 

 Recognizable bacterial remains, however, are stated to be extremely 

 short-lived and it is difficult to identify them even in modern bog- 

 iron-ore deposits. There is also to be considered the fact that 

 nowhere in the ocean do we know of extensive deposits of ferric 

 hydroxide being formed at the present time. This objection, 

 however, is equally valid in case of ferric hydroxide precipitated 

 either chemically or biochemically, and therefore must be con- 

 sidered under either hypothesis. In the formation of bog ores on 

 continental areas bacteria are known to play an important part, 

 but chemical precipitation is probably effective also, and it is 

 difficult to say which plays the principal role. In spite of various 

 uncertain factors, however, the deposition by micro-organisms 

 seems more easily to explain the various unusual phases of the 

 Brazilian iron-formation sedimentation, and we prefer to adopt 

 this hypothesis, at the same time, however, realizing fully the 

 inadequacy of our present knowledge concerning bacteria as a 



