»62 DE. A. HOLMES ON THE PRE-CAMBEIAN [vol. lxxiv, 



The biotite-gneiss {No. 96) of the Norray district is a banded 

 variet}? - rich in brownish-green biotite, and containing a considerable 

 amount of oligoclase. Quartz and microcline are the remaining 

 essential constituents, and, seen in thin section between crossed 

 nicols, these minerals exhibit extreme granulation along certain 

 bands, oligoclase having suffered to a much smaller extent. Asso- 

 ciated with this gneiss are garnetiferous rocks of three distinctly 

 different types. 



No. 97 is a garnetiferous biotite-gneiss differing mineralogically from the 

 biotite-gneiss described above only in the presence of crystals (6 to 8 mm. in 

 diameter) of pale-pink garnet, and of a plagioclase that has the com- 

 position of calcic andesine instead of oligoclase. Wisps of biotite are 

 frequently moulded on the periphery of the garnets, and rounded crystals of 

 biotite are sometimes present as inclusions. It is thus possible that biotite 

 may have contributed the pyrope and almandine molecules of the garnet. 

 A qualitative analysis of the latter indicated the presence of magnesia, iron, 

 .and also lime. In the absence of definite field evidence, it is unfortunately 

 not possible to discuss further the origin of the garnet. The rock, as a whole, 

 clearly differs chemically from the normal biotite-gneiss in being richer in 

 lime-bearing minerals. (PL IX, fig. 4.) 



The next rock to be described from this region is a dark and heavy skarn- 

 like rock (No. 78) composed of pale-brown garnet and bright-green pyroxene 

 with interstitial scapolite (near to meionite) and quartz. As in the scapolite 

 found in the other garnetiferous rocks of Mozambique, the mineral contains 

 Tounded inclusions of zircon. (PL IX, fig. 6.) 



Finally, a beautifully foliated eclogite (No. 77) is represented, composed of 

 pale-pink garnet (wine-colour in the hand-specimen), green amphibole, and 

 colourless pyroxene. Entile is present as inclusions in all three minerals. 

 {PL IX, fig. 5.) 



The only other garnetiferous rock brought back l is a garnet- 

 bearing gneiss collected on the old military road from Fort Chinga 

 towards Mount Mushima. This rock contains alternating bands 

 rich in quartz and microcline respectively, grains of deep-red garnet 

 (pink in thin section) being distributed through both, though 

 more abundantly in the quartz-rich bands, while biotite is confined 

 almost entirely to the felspathic bands. In the neighbourhood of 

 the garnets, magnetite and apatite are plentiful, together with 

 occasional rounded grains of zircon, and the same accessories are 

 present as inclusions. A curious feature of the garnets is the 

 frequent presence around them of a narrow selvage of orthoclase. 

 The garnets are frequently oval in shape, or embayed by quartz 

 individuals, so that they send out stumpy arms with blunt ter- 

 minations. The long axes of the oval arm-shaped sections are 

 generally parallel to the foliation of the rock, and in these cases 

 the orthoclase rims first thicken out as they approach the ter- 

 minations and then come sharply to a point. This feature to some 

 extent supports the surmise hinted above, that garnet ma}' have 

 derived its ferromagnesian constituents from biotite, for such 

 a transformation would be accompanied by the liberation of the 

 constituents of orthoclase. 



1 Several others were collected by Mr. D, A. Wray, but these unfortunately 

 were lost on the homeward journey. 



