190 PROCEEDINGS OF THE ACADEMY OF [Part I 



The origin of the North CaroUna deposits has been a much dis- 

 cussed question. The contact zones were described early by She- 

 pard/2 Chatard," and others. 



Juhen,^^ although regarding the dunite beds as metamorphosed 

 sediments, believed that the veins originated by the introduction 

 of deep seated solutions of soda and alumina into fissures during 

 the period of metamorphism. 



Chatard pointed out the progressive increase in magnesia as the 

 dunite is approached and a gradual decrease in the amount of 

 alumina. He regarded the corundum veins as the result of altera- 

 tions along the contact of aluminous-alkaline silicate rocks and 

 magnesian-silicate rocks, probably effected by surface solutions. 



Because of the circumstance that many of the corundum deposits 

 occurred at the margins of peridotite masses, Pratt and Lewis*^ 

 presented the theory of their origin as magmatic segregations of a 

 peridotite magma. It was believed that the corundum was held 

 in solution in the molten peridotite when it was intruded, and that 

 the corundum was among the first minerals to crystaUize out. This 

 solidification would take place first at the outer border of the mass 

 where it cooled first. Convection currents would tend to bring a 

 new supply of material carrying alumina to this outer zone where 

 it would crystallize. 



Holland'*^, and TealP', presented objections to this theory based 

 on the rather old experiments of Morozewicz^*, which indicated 

 that in a magma supersaturated with silica, that is, one in which 

 the molecular ratio of AI2O3 to the bases CaO, KoO, and Na.O is 

 more than 1:1, the excess of silica would separate out as corundum 

 if MgO and FeO were absent and Si02 were not present in an 

 amount sufficient to form sillimanite. With the presence of MgO 

 and FeO spinel would form, or if enough silica were present iolite 

 (cordierite) . 



More recently the ternary system MgO-Al203-Si02 was studied 

 in detail by Rankin and Merwin.^^ The results of their precise 



42 Am. J. Sci. (3) 4: 109-114, 175-180, 1872. 

 "U. S. Geol. Surv. Bull., 42: 45-63, 1888. 

 « Proc. Boston Soc. Nat. Hist., 22: 141-149, 1882. 

 *^ Loc. cit. 



« Mem. Geol. Surv. India, 30: 209, 1901. 

 "Proc. Geol. Assoc, 16: 72, 1899. 



«Tschermak's Min. u. Petr. Mitth., 18: 1-90, 105-240, 1899. 

 *9 Am. J. Sci., (4) 45: 301-325, 1918. See also Olaf Andersen, Ibid., 39: 407- 

 454, 19^5. 



