REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 37 



Usually it takes the form of grains with rugose surface, sometimes broken, transparent, 

 with a blue tint inclining a little to violet, and with very decided relief ; in general 

 these sections are isotropic. Magnetic iron in the shape of grains or of microscopic 

 crystals is tolerably abundant, and is distributed throughout the mass. Lastly, there 

 are still to be noted some small prisms of apatite. 



The limestone mentioned above, which was collected by Mr. Buchanan in the south- 

 east of Rat Island, shows on a freshly-fractured surface a compact rosy or yellowish mass, 

 with small white crystalline specks and yellow or blackish grains less than a millimetre 

 in diameter. The white specks are shells, the yellow and black grains are fragments of 

 rocks and of volcanic minerals. This rock is moderately hard ; it often presents on its 

 surface a scoriaceous aspect, and sometimes also cavities are seen in the interior. 

 When treated with hydrochloric acid it leaves a residue of about 30 per cent, of its 

 mass. Under the microscope this rock resolves itself into crystallised colourless 

 limestone, devoid of any trace of organic structure, and forming, one may say, the 

 paste or cement of the clastic grains of organic or mineral origin. These grains of 

 calcite are of two sizes : some are very large, while others, smaller and probably of 

 secondary formation, occupy the intervals left between their larger neighbours. Car- 

 bonate of lime also occurs, in microscopic acicular crystals. The mineral and organic 

 particles are all clastic and worn, each being surrounded by a narrow zone of calcite. 

 Among the minerals olivine is frequently visible, coloured red by decomposition ; 

 other grains consist of small splinters of basaltic rock,— among them being some 

 particles of basaltic glass changed into palagonitic matter. A rock almost identical 

 with this limestone of Eat Island was found by Mr. Buchanan overlying the basalt of 

 Platform Island, of which we are about to speak. 



The islet of this group that goes by the name of Platform Island is composed of 

 columnar basalt, on which lies an extensive and uniform bed of calcareous rock, the 

 specimens of which are, as we have said, analogous to the limestone of Rat Island. 

 The basalt of Platform Island is sbghtly more granular than the nepheline basalt 

 described above. It is black, and slightly vesicular, while to the naked eye only some 

 crystals of augite are visible, embedded in the ground-mass. Under the microscope 

 this rock proves to be felspathic basalt. In a ground -mass, in which from their 

 number certain small felspathic lamellae predominate, we find large microporphyritic 

 crystals of augite, and sections of olivine of smaller size ; the plagioclase as well as the 

 magnetite are always microlithic. The augite sections are not very prismatic; the 

 crystals are more shortened than those usually observed in basaltic rocks. Sections 

 parallel to oopoo are often unsymmetrical hexagons, whose outlines represent the 

 traces of faces of the zone n and t, and of those of the prisms. Octagonal sections also 



