REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 169 



and hornblende. They present a fibrous structure such as the two minerals just named 

 do not possess in this rock ; the colour also is rather greyish, with a scarcely perceptible 

 red tinge. The sections are prismatic, with angular or rounded outlines, often very 

 irregular ; they give straight extinction, and some hexagonal sections remain dark 

 during a complete rotation between crossed nicols. The basal sections show polar 

 rings in convergent light. Pleochroism is not very pronounced, in fact one can hardly 

 detect any difference in tint. 



Magnetite is shown generally in octohedral crystals or in somewhat large grains, 

 but when these grains are without crystalline form it becomes difficult to say whether 

 the mineral is primary or whether the irregular sections were hornblende now replaced 

 by magnetite. 



Amongst the most interesting specimens collected at Camiguin we may mention, in 

 the first place, some fragments the mineralogical composition and texture of which are 

 altogether different from the andesitic volcanic products just described. The rocks 

 now under consideration are undoubtedly granitic, and they must be viewed as portions 

 of the underlying masses torn up and thrown out by the volcano. These inclusions 

 are instructive, because they show the deep modifications produced by the intense 

 caustic action of the volcanic magma in which they were embedded. To the naked 

 eye the specimens appear milky white, speckled with brilliant scales of black mica. 

 The white minerals have a vitreous aspect ; the constituent quartz and felspar which 

 compose this granular mass are not easily made out even with the lens. The rock 

 looks as if it were fritted, and crumbles readily into a powder of irregular grains like 

 those of pulverised glass or quartz. Microscopical examination reveals such decided 

 differences of composition and structure, between this rock and those of the volcano, 

 that it must be viewed as not belonging to the same formation as the andesites of the 

 Camiguin volcano, but should be classed with the rocks of granitic type. Thin slices 

 show a distinct granitoid structure in which monoclinic and triclinic felspars, quartz, 

 biotite, titaniferous iron, and minute augitic microliths take part. At the first glance it 

 is seen that some of the principal elements have not the microscopic appearance of the 

 minerals of a normal granite. They are corroded, cracked, full of gaseous inclusions, 

 and, what is in accordance with the principal features, a colourless amorphous material 

 is found infiltrated between the constituent minerals. This substance is perfectly 

 isotropic at the points where it is isolated, and it contains the characteristic crystals 

 which occur in the glassy cement of sandstones vitrified by contact with eruptive rocks. 

 In certain cases this glass appears to be cracked, and to be derived probably from the 

 fusion of the felspar. As the elements are almost never outlined by crystallographic 

 contours, and as they are deeply altered, specific determination is very difficult, 

 especially in the case of the plagioclases. Sections of these felspars are widely dis- 



(PHYS. CHEM. CHALL. EXP. — PART VII. — 1889.) 22 



