I08 i-r.NTTl KSKOI.A. M.-\. Kl. 



laliti rocks, and ihr .siig,t(e.stion seems to be supported by a striking- 

 chemical relationshi[) a[)parent from available analyses of the granulites'. 

 We may only quote the analysis (1) by Naima Sahlbom of a granulite- 

 gneiss from Finnish Lapland, composed of quartz, potash and soda feldspar, 

 biotitc, futilr and garnet, beside the figures for our garnet-ampjiib<')lite (III: 



This analysis of granulite-gneiss as well as the other analyses of 

 related rocks published by Hackman show all a very high proportion of 

 silica combined with a high percentage of iron in preponderance over 

 magnesia, just as the Kantalahti rock. It seems therefore most probable 

 that we have to do with a pétrographie province whose comagmatic rocks 

 are characterized by the chemical features just named. 



Before returning to the question whether the garnet-amphibolite of 

 Kantalahti may have originall}' been an eclogite we shall at first look at 

 its structural characters, illustrated by the microphotograph, pi. Ill, fig. 17. 

 Rounded grains of garnet, 8 — 10 mm in diameter, stained with minute 

 quartz-grains, are embedded in a foliated mass in which anhedral grains of 

 hornblende in poikiloblastic intergrowth with rounded enclosed small feldspar 

 grains form elongated lenticular groups in which also the grains of the 

 biotite, titanite and iron ore are concentrated. The granular quartz forms 

 similar groups, and the parallel arrangement of all these give the rock a 

 pronounced foliation. It is a typical structure of a crystalline schist. 



The rock could therefore very well, at some earlier stage, have been 

 an eclogite in which the pyroxene was latter converted into feldspar and 

 hornblende, the rutile into titanite, and the garnet alone been preserved as 

 a relic. A feature in the composition of the garnet would even support 



' \'ictor Hackman, Die chemische Beschaffenheit von Eruptivgesteinen Finnlands und der 

 Halbinsel Kola. Bull. Comm. géol. Fini. N:o 15, 1905, analyses 84 — 87. 



