A. Ilarlcer — Hocks from the Tonga Islands. 253 



are some which are highly calcareous. Specimens from the northern 

 part of Eua have a white chalky aspect, and effervesce freely with 

 acid, but they contain abundant felspar microlites and broken 

 felspar crystals, some giving nearly straight extinction (oligoclase- 

 andesine?), besides occasional enstatite, etc. 



A coarser type of rock, well represented in the collection, may 

 be termed a volcanic sandstone. It resembles an ordinary sandstone 

 of moderately coarse grain, though isolated fragments occur up to 

 about an eighth of an inch in diameter. On the generally yellowish 

 brown surface are seen little scattered broken crystals of black 

 augite and the lustrous cleavage-planes of felspar fragments. Some 

 of the felspar flakes give nearly straight extinction ; others are 

 evidently more basic. As before, the bulk of the fragments com- 

 posing the rock are of brown-stained glassy lava. Good examples 

 come from the mouth of Ana-ahu. 



At the last-named locality there is an alternation of harder and 

 softer beds. The softer are of the character just described ; the 

 harder differ from them in possessing a cement of calcite, which 

 forms a large part of the rock. These calcareous beds, which occur 

 also in other parts of the volcanic series of Eua, present under the 

 microscope some points of interest. One of these features, the 

 occasional presence of minute chips of characteristic metamorphic 

 minerals, has already been alluded to. Slices [1268, 1273] show 

 that the fragments, some of which appear rolled, are of brown- 

 stained andesite, enclosing many felspar microlites and occasionally 

 a porphyritic crystal. The larger pieces have the spongy character 

 of pumice. Besides the andesitic fragments, are seen little chips 

 of minerals such as might be derived from the same source ; clear 

 plagioclase, pseudomorphs of calcite after felspar, and brightly 

 polai'ising pyroxene ; also rarely a grain or two of quartz, seemingly 

 clastic. The matrix of crystalline calcite encloses numerous fora- 

 miniferal tests which seem to belong for the most part to Globigerina. 

 More rarely occurs a small chip of shell or a fragment recalling 

 the characteristic structure of an echinoderm plate. The calcareous 

 matrix further exhibits in parts a beautiful oolitic structure, giving 

 the black cross well defined in polarised light. The structure is of 

 the spherulitic type, with radial but not concentric arrangement, 

 and is quite different from the rolled oolitic grains well known in 

 our Carboniferous and Jui'assic limestones. It is produced in situ 

 by molecular action, as is sufficiently proved by the fact that one 

 of these spherulitic growths occupies each chamber of the forami- 

 niferal tests. 



It seems clear that the whole of these volcanic rocks of Eua are 

 of submarine origin. This appears from their uniform and horizontal 

 gtratification, the indications of the " sorting " auction of water, and 

 the occurrence at various horizons of highly calcareous rocks, some 

 containing marine organisms. Probably some parts of the deposits 

 come from the destruction of volcanic accumulations thrown up above 

 sea-level, such as Falcon Island at the present day, which is being 

 rapidly destroyed by the waves. 



