REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 25 



plagioclase are sometimes observed which are quadratic and show the striation of 

 albite and pericline. In these sections, which are in the zone P:k, the extinction 

 takes place almost parallel to the trace of M ; hexagonal sections extinguish at an 

 angle, the mean value of which does not exceed 6°. These measurements and the 

 nature of the rock appear to show that the plagioclase in question is oligoclase. All the 

 steps in the transition may be observed from rather large crystals of plagioclase to 

 the small entangled felspathic lamellae which form almost the entire ground-mass of the 

 rock. These microliths are polysynthetic and extinguish at very small angles, showing 

 that, from the point of view of the plagioclastic mixture, they are akin to the larger 

 individuals which belong to an earlier stage of consolidation. It is perhaps not without 

 interest to point out this analogy of the microliths of the base and the microporphyritic 

 crystals. The microscopic preparations are sprinkled with black grains of magnetite or 

 titaniferous iron. In addition epidote, and in particular calcite, may be mentioned as 

 rather common secondary products. Calcite occurs in somewhat large sections traversed 

 by polysynthetic lamellae following — ^ R. Finally, quartzose veinules were observed 

 penetrated by small colourless lamellae, which appear iridescent in polarised light, and 

 are very probably scales of white mica. On taking account of the facts that no trace of 

 calcite appears in the quartz veins, and no quartz in the sections of calcite, one is led to 

 conclude that the infiltration of quartz and of calcite occurred at different stages in the 

 series of secondary modifications to which this diorite has been subjected. 



In the two rocks just described the specimens were referred to diorite, and we 

 remarked the profound alteration which had attacked the hornblende in one case and 

 the plagioclase in the other. The difficulties in the way of exact determinations may 

 readily be understood when decomposition has to so great an extent veiled the true 

 nature of the rock, and when so many of the specimens are rolled pebbles picked up 

 on the shore. We incline to believe, however, that they belong to the ancient type, 

 and these general remarks apply equally well to the specimens from the same locality 

 which remain to be described. 



The next to consider is a fine-grained greenish rock, dotted with felspar, and 

 breaking with an irregular fracture. Microscopical examination shows the rock to be 

 greatly altered. The felspar is associated with secondary minerals, epidote, calcite, and 

 chlorite ; sections which might belong to bisilicates are not detected with certainty, but 

 everything goes to show that these were present in the rock before it was decomposed. 

 It is remarkable that the felspar should not have been more altered, the polysynthetic 

 lamellae being still perfectly apparent. These crystals are somewhat large, and appear 

 enclosed in a mass which is composed principally of minute lamellar sections of 

 plagioclase. These microporphyritic crystals sometimes present sections in the form of 

 an octagon with two long sides. This would indicate that the crystals have pyramidal 

 faces in the zone P : M (n) or in the zone x : M (o). The sections extinguish at a 



(PHTS. CHEM. CHALL. EXP. — PART VII. — 1889.) 4 



