ART. 2. PETROLOGY AT GOOSE CEEEK SHANNON. 31 



would occur at a comparative!}- early stage of cooling in the more dlfBcultly 

 fusible siliceous solvent. 



The basaltic magma, on the other hand, with its low melting point and it? 

 high content of dissolved basic constituents, would reach the point of saturation 

 for some of these (magnetite and olivine, for instance) at comparatively high 

 temperatures and while the lava is still quite fluid. If such minerals crystallize 

 any considerable length of time before the other constituents, the magma remain- 

 ing liquid, their concentration in the lower parts of the mass by gravitation 

 must result as a mechanical necessity, unless there are eddies or other currents 

 sufficiently strong to prevent; and such currents would probably prevent 

 differentiation by any process in the parts affected. In many rocks the ore 

 grains are much smaller than the silicate minerals, and would therefore offer 

 greater resistance to settling through the magma. In such cases gravitation 

 would affect the larger olivines particularly. 



In the next stage of crystallization, there would undoubtedly be the same 

 tendency for the augite crystals to sink and the feldspars to rise toward the top 

 of the sheet, but by this time the increasing viscosity of the magma and the 

 clouds of new minerals forming would doubtless prevent any extensive segregation 

 of these bj' gravitation. 



The degree of concentration finally attained by this process would depend 

 on the fluidity of the magma and the time intervening between the formation 

 of the first minerals and the next succeeding stages of crystallization. Further 

 the position reached by such descending minerals would be determined by the vis- 

 cosity of the magma toward its lower contact, that is, by the extent of cooling due 

 to the rocks into which it was intruded. 



The basic concentration forming the olivine-diabase ledge in the Palisades 

 was not formed at the cooler contact, nor is it duplicated in the corresponding 

 upper portions of the sill. Its formation can not, therefore, be attributed to the 

 action of Soret's principle or any other process of concentration due to cooling. 

 If regarded as the result of chemical differentiation before intrusion, it must be 

 an earlier or later injection than the accompanying diabase above and below, 

 but its uniformly coarse texture and its great regularity in thickness and posi- 

 tion with reference to the base of the sill would seem to preclude this hypothe- 

 sis. The great overlying body of diabase, however, has been entirely freed from 

 olivine, except at the upper contact, and this mineral has been lodged in the 

 remarkably distinct zone of olivine-diabase 10 to 20 feet in thickness and lying 

 40 to 50 feet above the base of the sill. The bulk of the diabase, however, is 

 somewhat quartzose, but it often passes into normal diabase, and toward the 

 contacts into a somewhat olivinic facies, which is more basic in character, 

 though much less so than the diabase ledge referred to above. 



The above is the only detailed discussion of differentiation of the 

 diabasic magma occupying sills in the Newark series known to the 

 writer, and is quoted because of the similiarity of the Palisade sill 

 to that of Goose Creek and because it forms an excellent exposition 

 of the principle of gravitative differentiation to serve as a back- 

 ground for other quotations. 



As compared with the above hypothesis of purely gravitative proc- 

 esses operating in a sill chamber of molten magma is the explanation 

 of Daly 12 of the Moyie sills of British Columbia, where large sills of 



"R. A. Daly, Geol. of the N". A. cordil!era at the 49th Parallel, Mem. 38, Can. Qeol. Surv , pp 226- 

 256, 1912. 



