64 Pratt — Origin of the Corundum associated with 



the dunite side of the vein and but a few on the gneiss side. 

 The decomposed material that would be found on the gneiss 

 side of the vein would vary according to the amount of the 

 dunite that had been held between the corundum and the 

 gneiss (fig. 8). The common decomposition product surround- 

 ing the corundum is the chlorite (clinochlore) or a further 

 alteration of the mineral to the vermiculitcs. 



Between the corundum imbedded in the chlorite and the 

 gneiss there is often but very little and at times no chlorite or 

 vermiculites developed, the corundum-bearing portion of the 

 zone being apparently in direct contact with the gneiss (b of 

 cross-section, p. 58). 



In the dunite veins the alteration products developed are 

 the same on both sides of the corundum-bearing zone and are 

 in most cases nearly an exact reproduction of the dunite side 

 of a cross section of a contact-vein (figs. 5 and 6). This de- 

 monstrates that the gneiss had no influence in the formation 

 of the alteration products of the contact-veins. 



The analyses of Chatard* which show the chemical character 

 of the vein to increase in magnesia and decrease in alumina as 

 the dunite is approached, are in accord with the present theory 

 regarding the formation of the alteration products. 



The penetration of molten material carrying corundum into 

 the gneiss during the differentiation of the molten dunite, 

 would explain the occurrence of any corundum found in the 

 gneiss bordering a contact vein. There is a similar occurrence 

 of the sulphides penetrating into the gneiss during the cool- 

 ing and differentiation of a molten gabbro, which is well 

 illustrated in Vogt's* article and have been reproduced by 

 Adams.f 



Pirsson'sJ theory regarding the origin of the Montana sap- 

 phires, occurring in a basic igneous dike,§ is that these crys- 

 stals separated out from the molten magma as this began to 

 cool. He points out that originally the rock could not have 

 been sufficiently rich in alumina to have allowed a general 

 separation out of that material, and undoubtedly the magma 

 took up quantities of inclusions from the sediments through 

 which it passed. Clay shales, or a rock of similar composition, 

 must have been among these sediments ; and the fragments 

 of this shale would be dissolved by the basic magma from 

 which on cooling the alumina would separate out as corundum. 



Summary. — This theory of the igneous origin of the corun- 

 dum associated with the peridotites is in accord with the field 



* U. S. Geol. Survey, Bulletin 42, pp. 50-56. 



f Already cited. 



% This Journal IV, iv, p. 422, 1897. 



§ This Journal III, 1, p. 467, 1895. 



