Cole — On the Growth of Crystals. 



119 



it were, is imparted to the material that is capable of crystallising as 

 amphibole ; and the mineral develops on a far bolder scale than was 

 exhibited in the original rock. Something of the kind is traceable in 

 composite specimens from the mass at Glendalough in County Wicklow, 

 although in the end an excessive development of biotite has set in. The 

 hornblende was for a time regenerated, but tended rather towards 

 actinolite; and the larger crystals are now partly replaced by a multi- 

 tude of flakes of mica. Dr. F. Katzer^ has recently noted in a Bohemian 

 instance how pegmatite dykes seem more common in amphibolites 

 that are rich in garnet than in those that are poor in garnet, and how 

 their junctions are very commonly accompanied by hornblende crystals 

 of exceptional size. The first remark reminds one of the frequent 

 association of garnet-am phibolites with granite in north-west Ireland. 

 In these cases the presence of the garnets seems due to the influence 

 of the intrusive mass on a mixed aphanitic and sedimentary series. 

 Garnet-amphibolite, from this point of view, is practically always a 

 product of contact-metamorphism ; and the more abundant the veins 

 of pegmatite or granite, the more garnet arises in the invaded mass. 

 Garnet has certainly developed, in sharply-edged rhombic dodecahedra, 

 in the altered amphibolite of Glendalough, and arises very possibly 

 from associated patches of schist ; while at Castleore, south of Lough 

 Gill, I have lately found that the enveloped amphibolite- becomes 

 very notably coarser when seriously attacked by granite. The crystals 

 of garnet, as well as of hornblende, run together, it would seem, under 

 the stimulating influence of the granite, producing a rock of very 

 conspicuous grain. 



Near the mouth of the Gweebarra in western Donegal, the granite 

 magma has invaded certain amphibolites of the " Dalradian " series, 

 without producing garnet. This fact in itself supports the suggestion 

 that an admixture of aluminous sedimentary material is requisite for 

 the development of ordinary eclogite and garnet -amphibolite by 

 contact-action. The interest of the Gweebarra cases lies, however, in 

 the coarseness of the resulting quartz-amphibole-diorite in com- 

 parison with the grain of the diorite or amphibolite that has been 

 attacked. 



On Sheet 15 of the Geological Survey Map of Ireland, two dykes 



1 "Die Magneteisenerzlagerstatten von Maleschau und Hammerstadt," Ver- 

 handl. d. k. k. geol. Keichsanstalt, 1904, p. 199. 



- This contact was originally described in Proc, R. I. Acad., vol. xxiv., sect. B, 

 p. 364. 



