REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 117 



basaltic lapilli cemented by a palagouitic matter. The microscope shows that this 

 zone of contact, which resembles tachylite, is essentially composed of a vitreous base 

 containing olivine and small rhombic tables of plagioclase, similar to those just referred 

 to as occurring in the palagonitic tufas. The vitreous part, resulting from the rapid 

 cooling of the eruptive rock in contact with the surrounding mass, can be observed in 

 the microscopic preparations joined to the rock forming the central part of the vein. 

 This more crystalline zone is composed of the same minerals ; the plagioclase crystals, 

 however, take another form : instead of the tabular sections just referred to, they are 

 prismatic, and often in the shape of skeletons forked at two extremities. Augite is not 

 developed in it, but the brownish glass is darker, and it is filled with trichites and 

 spherulites. Olivine often occurs in twinned crystals, which are sometimes sharply 

 outlined by crystallographic lines in one part of the section, and in the other part shade 

 off into worn and irregular forms. The large sections of olivine in this rock are often 

 enclosed in felspathic lamellae. On the other hand, the felspathic microliths are 

 surrounded by sections of olivine, which, from this point of view, seems to play the 

 same part as augite does in many basalts. To return for a moment to the rhombic 

 tables of plagioclase, which are confined to the vitreous zone in contact with the 

 surrounding rock. It is natural to suppose that the development of these tabular 

 crystals is in relation with a particular state of consistence of the lava where they were 

 formed. These tabular crystals of plagioclase show the faces P and x, and sometimes 

 those of y. The angle of extinction measured on the face M is negative, and about 32°. 

 This observation suffices to show that this felspar is allied to bytownite. 



The coal-beds of this part of the island are associated with schistoid rocks, which 

 resemble certain slaty rocks. At first sight one would mistake them for slates of 

 slight fissility. Their colour is purplish, and the surface shines like some clays, but 

 they are harder, and the streak is not lustrous. No constituents can be discerned by 

 the naked eye. The microscope proves their volcanic nature, and that they belong to 

 the eruptions which poured out trachytic lavas. In ordinary light small prisms of 

 augite and grains of magnetite are seen in a colourless and homogeneous ground-mass. 

 With polarised light a rather large number of sanidine sections are seen in the prepara- 

 tion. They sometimes assume the form of elongated lamellae, but they are generally 

 placed with their widest faces parallel to the cleavage of the rock. Sections parallel to 

 M often show the Carlsbad twin with k as the plane of composition ; sometimes the two 

 twinned individuals are not entirely superposed over the whole extent of the face M. 

 The crystals are generally broken up, and present undulating extinction, induced by the 

 mechanical strain to which the schistose character of the rock is also due. The whole 

 mass seems to have been penetrated by chalcedony. 



The hills situated to the north of Christmas Harbour, and reaching an altitude of 



