THE DECCAN TRAPS 773 



spend with what might be expected on the complete crystallization of the 

 vitreous portion of a Deccan or Gregonian basalt. 



The crystallization of magnetite in the final stage in much gabbro, 

 norite, and anorthosite (contrary to Eosenbnsch's dictum) is discussed 

 recently by Vogt, 10 and he points out that a portion of the magnetite is 

 one of the last minerals to form, even in granitic rocks. The period of 

 crystallization of the magnetite is determined very largely by its per- 

 centage amount in the magma; so that if much of it be present, some 

 will crystallize out first, whereas it will be last in rocks poor in magnetite. 



So far as I can ascertain from the literature, only four analyses of 

 Deccan traps have been published, 11 none of which are satisfactory. I 

 have made 11 analyses of the specimens sent me from the India Survey, 

 endeavoring to obtain a fair idea of the general chemical characters of 

 the whole range of the Deccan traps by analyzing specimens which repre- 

 sented the various textural types and which also represented widely sepa- 

 rated portions of the vast area. Many more are doubtless needed to do 

 the region full justice; yet those made will serve to show the general 

 , features, especially as there seems to be considerable uniformity among 

 these basalts. The results are presented in Table I. 



The most striking feature of this series of analyses is the uniformity 

 in composition of the majority of the basalts, eight of them varying in 

 silica only from 48.6 to 50.1 per cent, while there are three with higher 

 silica, from 53 to 53.46 per cent. The larger group resembles most 

 basalts in all the main features, except that the iron oxides are decidedly 

 high, varying from about 12.6 to 14.5; in the average, 13.1, ferrous oxide 

 dominating greatly over ferric in all. The more silicic group does not 

 show such high iron oxides as is to be expected, but in these also the 

 ferrous oxide is much the higher. Corresponding to this FeO, the 

 amount of MgO is low ; so that the augite evidently contains much of the 

 hedenbergite molecule. Potash is notably higher in the more silicic 

 basalts. The CaO is about that which is normal to basalts. Titanium 

 and phosphorus are both in about normal amount, as is manganese, but 

 the absence of chromium is worth noting. Another noteworthy feature 

 of these analyses is the uniformly high water, which applies both to that 

 above and below 110°, although, unexpectedly enough, it is not highest 

 in the more glassy forms. All of the rocks which were analyzed were 

 apparently fresh, so that the presence of water can not be ascribed to 

 weathering. 



10 J. H. L. Vogt : Jour. Geol., vol. 29, 1921, pp. 629, 634. 



11 "Bombite" (byalitic basalt?), Bombay, Dufrenoy, Mineralogie, vol. 4, 1859, p. 289; 

 Traps, Bombay Island, Wynne, Mem. Geol. Survey of India, vol. 5, 1864, p. 196; Basalt, 

 Poonab, Wartb, Geol. Mag. (v), vol. 2, 1905, p. 21. 



