544 MR G. W. TYRRELL. 



difficult. From the imperfect determinations possible it appears that the range of 

 composition is from AD45A1155 to AD75A1125, but the bulk of the felspar is probably 

 near the latter composition. In thin section the felspars have a bluish tinge owing 

 to the abundance of minute, dust-like inclusions. They appear to have suffered 

 magmatic corrosion, as their angles are rounded off, and they have been worn into 

 curious irregular shapes. 



The next most abundant phenocryst is a deep-green, strongly pleochroic horn- 

 blende forming irregular plates which poikilitically enclose numerous rounded grains 

 of quartz. Frequently the crystals contain a core consisting of irregular grains of 

 quartz and magnetite, devoid of the green hornblende. The plates frequently retain 

 optical continuity, but occasionally they have been broken up into individual rounded 

 grains intermixed with the quartz mosaic. Next to the hornblende in abundance are 

 the magnetite-cassiterite intergrowths. These generally consist of a highly irregular 

 core of magnetite invested by a thin sheath of cassiterite. The latter mineral is 

 easily recognised by its uniaxial character, its well-marked twin-lamellation, and 

 decided pleochroism from yellow to reddish -brown. 



Biotite is evenly spread as minute anhedral shreds amidst the quartz mosaic, but 

 is more prominent as a constituent of curious clots or felts of the coloured minerals, 

 which are a feature of the rock. These clots consist of the magnetite-cassiterite 

 intergrowths, with numerous flakes of biotite, and the poikilitic hornblende, which 

 occasionally envelops the other constituents. 



The groundmass consists of an even mosaic of quartz grains with subordinate 

 water-clear untwinned felspar. The whole aspect of this rock, as described above, is 

 suggestive of recrystallisation at a not very high temperature. It is clearly a 

 hornfelsed porphyry belonging to the granodiorites, although quartz does not occur 

 among the phenocrysts. 



A rock from N. of Huambo (148) differs from the above in several particulars. 

 Amongst the phenocrysts quartz is abundant, and orthoclase is at least as abundant 

 as oligoclase. Hornblende is absent, but large flakes of muscovite enter as a con- 

 stituent of very closely felted clots consisting mostly of biotite flakes. The ground- 

 mass is of the same nature as that of the Bailundo rock, but is of much denser texture. 

 This rock is clearly a granite-porphyry. In another specimen from the same locality 

 (145) the secondary biotite is very abundant, and epidote enters the biotite felts as a 

 prominent constituent. From 307^ kilometres on the Benguella Railway (124) comes 

 a .specimen of the same nature, but the biotite forms wavy, sub-parallel streaks, 

 which, in thin section, give a semi-foliated aspect to the rock. 



A rock from N. of Cruz's (233) clearly belongs to the same group as the above. 

 It contains, however, no quartz phenocrysts, and only a few of felspar. The latter 

 are highly altered and are now little more than masses of granular epidote. The 

 new-formed biotite is not aggregated into clots, but is scattered evenly over the 

 i|iiartzose groundmass. 



