Trans. N. V. Ac. Set. 14 Oct. 23, 



The most important deposits of iron ore, known to exist in the 

 United States, may be grouped as follows : 



1. Magnetic ores of the Laurentian rocks of Canada, the Adiron- 

 dacks, and the Alleghany belt. 



2. The Huronian hematites of Lake Superior and Missouri. 



3. The crystalline ores of the Rocky Mountains and the Wasatch. 



4. The limonite ores of the Atlantic slope and the Mississippi Valley. 



5. The Clinton ores. 



6. The carbonates of the Coal-Measures. 



7. The spathic carbonates of New England, Idaho, etc. 



The Laurentian magnetites form lenticular sheets, sometimes more 

 than one hundred feet in thickness, and extending half a mile or more. 

 In many instances they are enclosed in walls of gneiss, slate, or marble. 

 They often contain much titanium, and have, as almost universal impuri- 

 ties (somewhat vicarious with titanium,) pyrite and apatite. Some of 

 the beds are also highly charged with manganese ; they usually contain 

 but little silica and alumina, and are the richest of our iron ores. 



The characters they present, which seem incompatible with the theory 

 of mechanical accumulation, are the following. First, their great magni- 

 tude, combined with their prevailing purity. Secondly, the large quantity 

 of apatite contained, which we must regard as of organic origin and there- 

 fore supporting the chemical theory, and so different from the magne- 

 tite in gravity as impossible to be mingled with it by any mechanical 

 agency. Thirdly, the occurrence ot great sheets of magnetite between 

 strata of limestone, like the Franklinite ore, or enclosed in layers of 

 limestone and argillite, like the bog-bed at Marmora, Canada. The 

 fact, that the slate and the limestone are deposits from deep and quiet 

 water, seems incompatible with the view that these strata have been 

 brought into their present association by mechanical means. Fourthly, 

 the frequent impregnation of magnetites with manganese, a material of 

 much lighter gravity, and an almost constant constituent of limonite, 

 goes far to prove such magnetic beds chemical rather than mechanical 

 deposits. Fifthly, the aluminous magnetites, like that of Croton Land- 

 ing, which contain almost no silica, could hardly have been formed by the 

 agency of shore waves, which always mingle more or less sand with 

 whatever they deposit. 



For these reasons, I must consider the mechanical theory as inade- 

 quate to account for the genesis of most of the magnetites of Canada 

 and the Eastern States. 



The specular ores of Lake Superior and Missouri offer objections to 

 the mechanical theory, in both their regularity of deposition, and their 

 mineral character. First, ihey are sometimes interstratified with jasper 

 — once fine silicious sand — in layers of great regularity and of extreme 



