302 J A. THOMSON. 



The faggot-like areas give an aggregate polarisation 

 colours like those of carbonates, but under high magnifica- 

 tion this is seen to be due to the presence of numerous 

 minute flakes of biotite. The small rods appear most often 

 to be opaque, but on the edges of the section they are seen 

 to consist of the green mineral described above. Finally 

 there are a few elliptical areas like amygdules, consisting 

 of coarse grained quartz and large flakes of biotite. 



The interpretation of such a rock is impossible without 

 field work to fix its geological nature, and in case it is an 

 alteration product, to enable one to trace stages from some 

 recognisable form. Assuming the small rods to be pseudo- 

 morphous after hornblende, the rock might represent a fine 

 grained amphibolite or more likely a camptonite such as 

 that described above. Such an assumption is, however, 

 little removed from guesswork. The rock is certainly not 

 a tachylite. 



The remaining ten rocks, though not all from the same 

 locality, form a distinct and related group. They show a 

 graduated mineral composition, and differ chiefly in grain 

 size. They are remarkable for the freshness of their 

 felspars and pyroxene, and although the olivine and iron 

 ores are at times somewhat altered, there is no sign of 

 saussuritic, sericitic, chloritic or epidotic alteration pro- 

 ducts. They must, therefore, be assumed to be of much later 

 age than the gneissose, and probably also the dyke rocks, 

 just described. They are for the most part noncrystalline, 

 coarse grained and almost black, the felspars being so 

 transparent as to affect the colour of the rock very little. 

 The mineral composition is that of gabbros or norites, but 

 on account of the perfect ophitic structure displayed, the 

 term dolerite is preferable. The more basic rocks contain 

 much olivine, the most acid contain free quartz in micro- 

 pegmatite, while the whole series is characterised by the 



