the Peridotites in Worth Carolina. 63 



dum and the dunite, in others there would be a somewhat 

 gradual passage from the corundum to the pure dunite, as 

 illustrated in fig. 8. 



The pressure and the temperature and other physical condi- 

 tions would affect the crystallizing and the separating out of 

 the minerals from the molten mass, and this will explain the 

 great variations observed in the corundum found in the same 

 mass of dunite. Thus, at Corundum Hill there occur masses 

 of block corundum, large rough crystals, small well-developed 

 crystals, and particles or grains of corundum. 



The differentiation of the basic dunite magma upon cooling 

 is similar to what Yogt- and Adamsf have described concern- 

 ing the separation of sulphide deposits from a molten gabbro 

 magma. The segregation or concentration of these ores is 

 usually near the contact of the gabbro with the gneiss and is 

 always sharply separated from the gneiss. While there is 

 sometimes a distinct line of separation between the ore and 

 the gabbro, at others there is a gradual transition from the ore 

 to the pure gabbro. It might be expected that the corundum 

 would be found surrounded by the dunite, but as was stated 

 on page 55, there is but one locality known to the author 

 where the corundum is found directly in the dunite. The 

 absence of any general occurrence of the corundum in dunite 

 is readily explained by the ease with which the dunite itself 

 suffers decomposition. 



The corundum concentrated in the dunite near the contact 

 with the gneiss, where the thermal waters coming in contact 

 with the heated masses would have the best chance to act, 

 would furnish alumina for reacting with the dunite to form 

 the aluminum-magnesium silicates found surrounding the 

 corundum. That the dunite decomposes very readily is appar- 

 ent from the numerous specimens found of the dunite com- 

 pletely surrounded by foreign material that must have been 

 formed at its expense. The corundum has been found in the 

 serpentine but this is often surrounded by chlorite. 



The descriptions of the cross sections of the corundum 

 veins given on pages 57-60 are similar to those described by 

 Shepard^:, Chatardg and Lewis ||, and show practically the same 

 sequence. 



As the dunite and corundum began to alter and decompose 

 by the action of atmospheric agencies and thermal solutions, 

 there would be formed a series of decomposition minerals on 



* Zeitschr. fur Prak. Geol., Nos. 1, 4 and 7, 1893. 



f On the Igneous Origin of certain Ore Deposits ; read before the Gen. MiD. 

 Asso. of the Prov. of Quebec, Montreal, Jan. 12, 1894. 

 X This Journal, III, iv, p. Ill, 1872. 

 § U. S. Geol. Survey, Bulletin 42, p. 49, 1887. 

 N. C. Geol. Survey, Bulletin 11, p. 93, 1896. 



