66 DR. A. HOLMES ON THE PEE-CAMBRIAX [vol. lxxiV,. 



large areas of country are low, pointing to a lateral flow of the 

 magmas from elongated regions of upward flow, and to the 

 existence of a considerable pressure from above, which can have 

 been caused only by the weight of superincumbent formations. 

 The dips and strikes of the foliation -planes thus lead us to the 

 conception of lateral interpenetration of schistose formations by- 

 granitic magma, fed from elongated regions of dominant upward 

 intrusion. 



However, the parallel structure of the gneisses may perhaps be^ 

 otherwise explicable. It is possible that biotite-crystals, formed at 

 an early stage in a viscous flowing magma, would be to some 

 extent segregated and drawn out into long streaks, between which 

 the later minerals would be obliged to crystallize. The richness 

 of the gneisses in biotite would thus correspond with the more 

 mafic character of the portions of the magma first to crystallize, 

 as compared with the later parts, now represented by the leucocratic 

 gneissose granites of some of the inselberg peaks. In addition,, 

 the later parts of the magmas generally are relatively enriched in 

 ' mineralizing agents ' or volatile fluxes, 1 and it is with the rocks the 

 formation of which has been controlled by such fluxes that the more 

 radioactive minerals are usually associated. 2 It might therefore 

 be anticipated that the leucocratic facies of the gneiss should be 

 richer in radium than the biotite-rich facies. Opposed to this- 

 view, which explains the structures by differentiation and viscous 

 flow, is the fact that the granitic cores show no signs of having 

 crystallized under the special influence of volatile fluxes. The 

 textures are granulitic, and the minerals, except for porphyritic 

 crystals, are small, and give the impression of having crystallized 

 in a stiffening viscous magma that flowed with difficulty. 

 Pegmatites, when present, are always of later age. Moreover, the 

 hypothesis fails to throw any light on the peculiar association of 

 the crystalline limestones with the gneisses. 



Considering all the evidence and the probabilities of the case, it 

 is thought possible that — 



(1) the gneisses were produced by the concordant injection of granitic 



magma into a series of pre-existing sediments or their metamorphosed 

 equivalents ; 



(2) the greater part of these sediments, being argillaceous, became 



granitized with such ease that, apart from the richness in biotite of 

 some of the gneissic bands, they have left no recognizable traces 

 of their former existence ; 



(3) when limestones or dolomites were encountered, the granite was partly 



desilicated, and formed hornblende-gneisses, amphibolites, and garnet- 

 iferous rocks, and in many instances these rocks preserved a core 

 of crystalline limestone from the further attack of the magma ; 



(4) the gneisses may be interpreted as composite rocks in which, as a rule,. 



the granitic element predominates. 



1 On the other hand, C. N. Fenner (Journ. Geol. Chicago, vol. xxii, 1914, 

 p. 594) believes that the mechanism of lit-par-lit injection depends on the 

 fluxing power of pneumato lytic gases which go in advance of the main, 

 granite magma, and so prepare the way for its penetration. 



2 A. Holmes, ' Science Progress ' 1914, No. 33, p. 19. 



