MACROBERTSON LAND AND KEMP LAND 183 



abundant constituent in the associations already indicated close to the selvedge of coarse 

 structure (cf. Plate XXXVIII, figs. 2, 3 and 4). 



Field data available concerning the rocks under discussion do not permit unequivocal 

 decision whether they are to be regarded as bands forming dykes in the acid gneisses or 

 band-like inclusions of earlier date. The micro-structures and mineralogy indicate clearly, 

 however, that they are metamorphosed gabbros and there can be little doubt that they 

 have suffered transformations while they were essentially solid. 



The development of garnet as a corona mineral intervening between pyroxene and 

 felspar and in granulitic aggregates of pyroxene is a familiar phenomenon which has 

 been the subject of much study since Kemp 1 first described the garnetiferous gabbros 

 of Essex County in New York State. The literature contains discussions of the chemical 

 mechanism of garnet formation to which reference is now directed. 2 



The sporadic development of this mineral and its distribution in specimen 4 indicated 

 above suggest: 



(1) If these rocks are dykes then the changes produced, especially the development 

 of reaction rims, may be a deuteric transformation following closely upon consolidation 

 by the passage of liquid or fugitive constituents emanating from the magma chamber 

 from which the dykes were erupted. Such an origin has been ascribed to the garneti- 

 ferous gabbros intruding granites, etc., of the Adirondacks by Gillson and others. 3 



(2) An alternative hypothesis may be mentioned. If these bands are inclusions their 

 metamorphism may be ascribed to thermal metamorphism by the enclosing acid gneisses. 

 On this view the coarse hornblende-plagioclase ± quartz aggregates on the face of 

 specimen 4 would then represent a reaction zone between the basic inclusion and the 

 enclosing gneiss, analogous to the hornblende-pegmatites forming a halo to amphibolite 

 inclusions in the Laurentian gneiss of the Haliburton-Bancroft area 4 or in the Lincoln 

 gneisses of Eyre Peninsula, South Australia. 5 In either case the manifold environment 

 of garnet in the rocks under discussion necessitates the incoming of metasomatizing 

 fluids, for it is clear that garnet development cannot be wholly ascribed to chemical 

 reactions within a limited closed system. 



1 igth Ann. Report U.S. Geo]. Sarv. 1897-8, Pt 3, p. 400. 



2 (a) Austr. Antarctic Expedition Reports, Series A, 1918, Vol. ill, Pt 1, p. 168 et seq. (b) Quart. Journ. 

 Geol. Soc. London, 1921, lxxvii, p. 116. (c) B.A.N. Z.A.R.E. Reports, 1937, Vol. II, Pt 1, p. 13. 



3 Journ. Geol. 1928, xxxvi, p. 149. 



4 Memoir Geol. Surv. Canada, 1910, No. 6, p. 76. 



6 Quart. Journ. Geol. Soc. London, 1921, lxxvii, p. 83. 



