772 H. S. WASHINGTON DECCAN SALTS AND PLATEAU BASALTS 



served in basalts. What is of special interest;, however, in this connection 

 is the late stage of crystallization of the magnetite and of much of the 

 augite. The rarity of angite and the practical absence of magnetite as 

 inclusions in the feldspars, especially in those of ophitic types, as well as 

 the progressive rarity of well crystallized magnetite with increase in 

 glass content, together with a somewhat less well marked decrease in 

 augite as compared with feldspar as the rock becomes more glassy, point 

 to the conclusion that nearly the whole of the feldspar crystallized first, 

 followed by much of the augite (probably with a period of some simul- 

 taneous crystallization), while the magnetite belongs to the latest stage. 

 It would follow that the glass should have a composition corresponding 

 to a mixture of augite and magnetite. That the glass is hydrous is indi- 

 cated by the uniformly high content in water shown by the analyses, and 

 that the greater portion of this water is original and not brought in by 

 weathering is evident from the very fresh appearance of all the minerals. 

 It would seem to be probable, therefore, that if any differentiation by 

 settling of crystals has taken place in these Deccan flows, we would ex- 

 pect an uppermost layer rich in feldspar, but would not expect a lower- 

 most layer rich in augite or olivine. Only study of the flows in the field 

 can decide whether these expectations are realized or not. 



In connection with the late crystallization of augite and magnetite or 

 their replacement by glass, the banded gabbros of Skye at Druim an 

 Eidhne, first described by Geikie and Teall, s are significant and instruc- 

 tive. Here we have a gabbro mass made up of bands, some of highly 

 feldspathic gabbro, others entirely of augite and magnetite, and inter- 

 mediate bands with less feldspar and hornblende than those of gabbro, 

 but more augite and magnetite, the last frequently inclosing the other 

 minerals. The chemical relations are shown by three analyses, the light 

 gabbro bands having 52.8 per cent of silica, a dark intermediate one -L0.2, 

 and an augite-magnetite band 29.5. Here is a striking analogy witli 

 what is seen in the Deccan basalts (and also the Oregonian), except that 

 at Skye the plutonic conditions permitted the complete crystallization of 

 the augite-magnetite portion, which solidified as glass in the basalts. 

 One is also reminded of the noncrystalline arapahite (magnetite basalt) 

 of Colorado, 9 which is composed of 25 per cent of bytownite. 15 of augite, 

 56 of magnetite, and 1 of apatite. This peculiar magnetite basalt occur- 

 as blocks in a volcanic breccia, and its relations to the ordinary basalts 

 of the region are unknown; but its modal and chemical characters corre- 



8 Geikie and Teall : Quart. Jour. Geol. Soc. vol. 50, 1S94, pp. 643-659. See also 

 Barker: Rocks of Skye, pp. 117-121. 

 a Washington and Larsen : Jour. Wash. Acad. Sci., vol. 3, 1013, p. 440. 



