REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 151 



pseudomorphism is effected is as follows. The hornblende becomes darker in 

 colour, the pleochroism more intense, the polarisation tints approach to dark-red 

 tones, and the sections assume a lamellar texture, the lamella? appearing undulated on 

 the surface in polarised light. In fact we see all the characters of hornblende being 

 exchanged for those we are accustomed to associate with black mica, but the form of 

 the sections is unaltered. We shall show immediately, that biotite exists as a 

 primary mineral in the rocks of Kandavu, and must point out the peculiarities which 

 make it possible to distinguish this from the secondary product just described. In 

 some cases the form of the sections gives no assistance, because both hornblende and 

 black mica may appear in thin slices as hexagonal sections. Yet it is possible to 

 demonstrate the secondary origin of the biotite, for, when this is the case, its hexagonal 

 sections show lamellae parallel to one of the sides of the hexagon; an observation 

 sufficient to prove that the biotite is of secondary formation. A hexagonal section of 

 biotite could not present this appearance ; the lamellae would not show themselves, 

 and the section would appear uniform. Those lines which appear in the sections, and 

 are caused by the union of lamellae of biotite, cannot be mistaken for the cleavages of 

 hornblende. Even if the characters of the mica were not so clear, this supposition 

 could not be reconciled either with the outlines or with the direction of the supposed 

 cleavages. The observations tend to prove that the lamellae of biotite are piled up 

 parallel to one of the pinacoids of the hornblende. 



There is little to say of biotite as a primary mineral. At first sight it closely 

 resembles hornblende, being surrounded, like the latter, by a black opaque zone ; but 

 its pleochroism, its pronounced lamellar structure, its reddish polarisation colours, 

 its brilliant tints between crossed nicols, and the characteristic undulating shades on the 

 surface of the section, prevent one from confounding this mica with anything else. It 

 is recognised as a primary mineral by its sharp outlines, either hexagonal or in the 

 form of a parallelogram, and by its always appearing isolated in the ground-mass. 



Augite is rather uncommon ; some microporphyritic sections of the mineral are of 

 a green colour, such as it often assumes in andesites. Bronzite is of more common 

 occurrence than monoclinic pyroxene. Olivine appears only in one of the specimens 

 from Kandavu which were examined, where it is an accessory element. Its sections 

 were of the usual rhombic or hexagonal form with worn outlines. It is a hyalosiderite 

 converted into hematite, and full of trichites. 



One of the specimens from Kandavu is an augite-andesite. It is a coarse-grained 

 rock, showing to the naked eye a greyish paste, enclosing crystals of plagioclase, from 

 2 to 3 millimetres in diameter, and small grains of greenish augite, with a few points 

 of black hornblende. Under the microscope this rock differs from that previously 

 described by the predominance of a vitreous base and the presence of microporphyritic 

 crystals larger than those of the amphibolic andesite just mentioned. Hornblende 



