lOS ME. C. E. TILLEY OX THE GEAXITE-GXEISSES [vol. Ixxvii, 



serve as do reliable guide in tliis determination. There are. how- 

 ever, certain peculiarities of some of these amphibolites which 

 deserve further mention, and they will be now discussed. 



The develoj^ment of plagioclase and hornblende (usually with a 

 subordinate amount of quartz-minerals), of which in the main 

 these amphibolites are constituted, is to be regarded as the result 

 of metamorphism of igneous rocks of general gabbroid or doleritic 

 composition. The production of hornblende from an original 

 pyroxene is probably not a simple change. In the first place, the 

 amphibole has been totally recrystallized, and is not pseudomorphic 

 after the original pvi'oxene. Its formation can be expressed in the 

 simplest form by the following equation : — 



In this case the amphibole is shown to contain the Tschermak 

 silicate, the alumina in manr analvses of hornblende conforminsr 

 to this constitution. It is to be noted that the resulting amphi- 

 bole is more aluminous than the original pyroxene from which it 

 is derived, and this is in liarmony with, typical analyses of amphi- 

 bole and pvi'oxene, the pyi'oxenes of many basic igneous rocks 

 showing a comparatively low alumina-content. With the release 

 of lime from the pyroxene-molecule, silicate is set free, and 

 appears as quartz in the resulting amiDliibolite. In some amphi- 

 bolites this lime appears as titanite. and it is probable that the 

 pvi'oxene is the original source of the titanium, thus 

 Ca(;OH),^TiO,-hSiO, = CaTiSiO--fH^O. 



Obviously, of the silica released in the main reaction, only that 

 amount which is molecularly equivalent to the titanium not 

 absorbed in tlie newly-formed amphibole, is taken up in sphene. 

 The characteristic association of hornblende and sphene in some of 

 these amphibolites (especially in the garnet-amphiboHtes) is thus 

 to be explained. 



Of subsidiary reactions which may assist in amphibolization we 

 may note the following : — 



(1) 2MgSi03-rCa3Io-Si,0, ^CaMg^SLO^^ 

 and (2) CaMgSi,0,-f MgAl,SiO, = Ca3Io-^Al,(SiOj3, 



Of (1) it may be noted that any enstatite- or hypersthene- 

 molecules present in the original pyroxene substance would, with 

 the diopside-molecule, yield tremolite. It seems possible that this 

 reaction might take place in rocks containing individual hyper- 

 sthene and within the limits of diffusion of a neighbouring mono- 

 clinic pyroxene, thus affording an explanation of the absence of 

 anthophylhte in those amphibolites the analyses of which corre- 

 spond to hvpersthene-bearing dolerites. 



Lastly {2) may be taken to represent the change, where the 

 amphibole substance is represented as containing the syntagmatite- 

 molecule, which, according to E. Scharizer,^ is the case for manv 

 amphiboles. 



1 Xeues Jalirb. vol. ii (1884) p. 143. 



