Many of the iron and manganese deposits of the Appalachian region are supposed by many to be replacement deposits. N. S. Shaler[31] in 1877 suggested that some of the iron deposits of Kentucky and Ohio were formed by the solution of iron from certain rocks, and its deposition in the form of carbonates by replacement with underlying limestone. Subsequently it was changed by oxidation to brown hematite. A notable case of replacement has also been shown by R. D. Irving and C. R. Van Hise[32] in the iron deposits of the Penokee series of Michigan and Wisconsin. Here the ore is supposed to be partly a replacement of chert in a trough between quartzite and igneous rocks. The solution that contained the iron was derived from strata in the same series of rocks in which the iron was re-deposited and contained a certain amount of manganese. It is shown how the iron and manganese were more or less separated in the replacement process and that the separation was due to the difference in the oxidability of the carbonates as explained on page 363. R. A. F. PENROSE, JR. FOOTNOTES [13] The exact age of the iron and manganese deposits here referred to is, in some cases, a little uncertain. Some may be Cambrian, others Silurian, but the exact determination of the age of the horizon is not a part of the present discussion. The matter has been discussed by the writer in Geological Survey of Arkansas, 1890, Vol. I., pp. 376–379. [14] See Geological Survey of Arkansas, 1890, Vol. I., pp. 320–325. [15] See Geological Survey of Arkansas, 1890, Vol. I., pp. 456–457. [16] R. D. Irving and C. R. Van Hise, U. S. Geol. Survey, Tenth Ann. Report, 1888–1889, Vol. I, pp. 409–422. [17] Geol. Survey of Penn., Vol. II, 1858, p. 739. [18] The solutions may be precipitated, as already shown, either with or without admixture with mechanical sediments; and there are in nature all gradations from almost pure deposits of iron and manganese ore to beds of shale, sandstone, etc. stained with iron or manganese. Subsequent concentration frequently causes decided changes in the latter deposits (see p. 370). [19] It has been suggested by A. A. Julien (Proceed. Amer. Assoc. Adv. Sci., Vol. XXVIII., 1879, p. 356) that in some cases the carbonates of iron and manganese may be only the fixed residue of organic compounds of more complex form once in solution in surface waters. [20] This oxide is generally in the form of the peroxide or the sesquioxide in a more or less hydrous condition. [21] Jahrb. des Vereins f. Naturkunde in Herz. Nassau, Vol. VI., p. 160 (Bischof). [22] Marcasite has the same composition as pyrite, but differs in crystalline form.