540 NEW TOKK STATE MDSIOJM 



cases, with a gray carbonate of iron, in beds underlying it. These 

 ore bodies are wholly in the limestone or between the limestone and 

 the adjacent slate or schist formations, or they are in the latter, 

 and as a rule of occurrence they are found on or near the dividing 

 line between these formations. Near Fishkill and at Shenandoah 

 the deposits are at the border of the Potsdam sandstone and at 

 the foot of the Archa3an ridges. The existence of the carbonate 

 ore in the deeper parts of some of the mines and interstra tilled 

 with the limestones is suggestive of the origin of the oxide 

 (limonite) by the decomposition of the ferriferous beds through 

 oxidation and the agency of carbonated waters, and of the great 

 masses of colored clays, also, through the disintegration and decay 

 of the slaty rocks and more argillaceous limestone.* The lime- 

 stone of these valleys and these overlying slaty rocks have been 

 studied by Prof. Dana, and are referred by him to the Trenton 

 limestone and the Hudson river slate formations, f 



The ore occurs (1) in large masses, somewhat cellular, having 

 the interstices filled with clays or sandy earths, (2) in cavernous and 

 hollow "bombs," often with beautiful mammillaryor stalactitic 

 incrustations on the interior, and (3) in irregularly shaped, frag- 

 mentary masses, distributed unevenly through the ochreous clays 

 ("ochres ") and sandy earths. The more solid ore has to be broken 

 down by blasting ; in the more earthy parts of the deposit it can be 

 picked down and nearly all of the ore be sorted by hand. In 

 mining, pits are sunk and worked open, or drifts are cut from the 

 pit, horizontally into the ore, and much of it is won by under- 

 ground work. In this district nearly all of the ore is mined from 

 open pits ; and some of them have reached vertical depths of over 

 100 feet. The ore is commercially known as "rock ore" or 

 " lump ore," that which is sorted by hand, and " wash ore," which 

 is the residue after the earths and sands have been removed by 

 washing. The brown hematite ores of Dutchess and Columbia 

 counties vary considerably in their chemical composition, all con- 

 taining more or less silica, little or no sulphur, but are rarely low 

 enough in phosphorus to answer for Bessemer pig-iron manu- 

 facture. Although there have been many ore localities dis- 



* For a clear and concise statement of the origin of these ores see " Note on the making of 

 Limonite ore beds," by Prof. James D. Dana, in Am. J. ur. of Science (3), vol. XXVIII, pp. £98- 

 400. 



+ Am. Jour. Science (3), vol. XVII, pp. 375-388 and vol. XXIX, pp. 205 et seq. 



