Review of Recent Geological Literature. 425 
manganese ore, poor in iron oxide but rich in CaCOa and MgCO s . It 
also explains why the limestones and dolomites adjoining the ore are on 
the whole richer in manganese and poorer in iron than the iron ores. 
(6) These iron ores, like the recenl lake and bog ores, are marked as a 
whole, though with numerous exceptions, by a larger apatite or phos- 
phorus contenl than tin- associated schists and limestones. 
(7) Of the rare elements, il is singular that gold alone in the native 
state is occasionally found in a relatively large amount in the iron ore 
(i g. in the itabirite in Brazil, Carolina. Africa, and Sweden). This is 
easily explained by the reducing action of ferrous carbonate (instead of 
the sulphate as in the metallurgical practice) upon AuCl.,. Analogy. 
Native gold is frequently found in stalactites of manganese ore in Aus- 
tralia. 
(8) Finally, it may be remarked thai the lenticular shape of our mag- 
netite and hematite deposits is explained by their analogy with the ir- 
regularly shaped beds of lake ore in relatively deep lakes. 
Tin- sedimentary iron ore deposits occurring in Sweden and Norway 
may be divided into a series of sub-groups (with the Swedish names): 
"Torrsten," "Quicksten" with "Enbart gaende," "Blandsten," and 
"Manganmalme," which are distinguished in the order named by (a) a 
decrease in the silica content and an increase in the content of basic 
silicates and carbonates, as regards their slag-producing constituents. 
We also find, in general, (b) an increaseof manganese, and (c) a decrease 
Of apatite content. 
The chemical evidence is repeated so often that we must ascribe to it 
a theoretical significance. It apparently sustains the most intimate re- 
lation with the details of the succession of precipitates. Under normal 
processes of oxidation (2FeCOs + O), Fe2 Os, with relatively large 
amounts of Si02 and P2O5 but with very little M11, is separated, and 
there may thus result a "Torrsten" even oul of a solution rich in M11. 
( )n the other hand, where, for any reason whatever (e. g. bj the presence 
of organic matter), oxidation is prevented or delayed and precipitation 
can only be accomplished by the aid of an excess of carbonic acid, we 
have contemporaneous deposition of iron, of a large pari of the Mn con- 
tent, and also of some lime (and magnesia I carbonate. We obtain thus 
a manganiferous iron ore mingled with carbonates and basic silicates. 
Analogy: the primarily formed siderite deposits of the later formations 
are distinguished by a relatively larger Mn content than the primary 
limonite and hematite deposits in the same formations. 
Tin Twenty-First Annual Report of tin Geological and Natural History 
Survey of Minnesota, for tfu year 1892. 8vo, 171 pages; Minneapolis, 1893 
The report is divided into seven parts, as follows: I. Summary state- 
ment and comparative nomenclature, by the state geologist, X. II. 
Winched. Accompanying this is a table of pre-Silurian rocksof Min- 
nesota witli their equivalents in adjacent territory. This table will lie 
of value in giving a concise statement of the nomenclature used b\ the 
Minnesota survey, and in showing the harmoiiv and lack of harmony 
