Mr J. D. Dana on the Origin of Trap Minerals. 197 



chrysolite ; and if much iron is present, we might have, with 

 the lime and alumina, the mineral epidote. The experiments 

 of Berthier, Mitscherlich, and Rose, and the facts observed 

 among furnace slags, confirm what is here stated. 



But not to go back to a resolution of the fused minerals into 

 their elements, we may consider for a moment what changes 

 the minerals themselves might more directly undergo in the 

 process of fusion. 



Much of the mica in granite differs from feldspar, in con- 

 taining half the amount in silica in proportion to the bases — 

 the bases in each being alumina and potash or soda. The 

 change, then, in the conversion of the mica into feldspar, would 

 require an addition of silica, which might be derived from the 

 free quartz of granite. Other varieties of mica contain mag- 

 nesia, which would go towards the formation of some mineral 

 of the magnesian series. It is possible that trachytes and 

 porphyry have thus been made from granite ; but trap rocks 

 could not have been so derived, as they contain from 10 to 25 

 per cent, less of silica. 



Again, hornblende and augite are so nearly related, that 

 they have been considered by Rose the same mineral, the 

 different circumstances attending the cooling giving rise to 

 the few peculiarities presented. There can be no difficulty, 

 therefore, in deriving augite by fusion from hornblende rocks. 

 This, moreover, has been actually confirmed by experiment. 



Augite, by giving up half of its silica, and receiving addi- 

 tional magnesia in place of its lime, is reduced to chrysolite.* 

 The Gehlenite, nepheline, anorthite, and meionite of Vesu- 

 vius, contain, like scapolite, only 40 to 45 per cent, of silica 

 and a large proportion of lime ; and it is no improbable sup- 

 position, judging from the small amount of silica, and from 

 the lime present, that scapolite rock, or rather limestones 

 containing scapolite, may have contributed in part towards 

 the lavas of that region. The ejections of unaltered granu- 

 lar limestones, and many mineral species pertaining to such 

 beds, strongly support this view ; and it is no less sustained 

 by the fact, that in the Vesuvian basalts, Labradorite, which 



* The formula of augite is R» 8i« ; that of chrysolite, R» Si. 



