Speight. — Andesites of White Island. 299 



Volcanoes," p. 361). There does not appear to be any general resem- 

 blance in the character of the flows from widely separated points of this 

 north-east line, even from those points which are generally accepted as 

 belonging to it. There is a considerable difference between the augite- 

 hypersthene andesites and the augite-hornblende-hypersthene andesites of 

 the Ruapehu group and the augite-olivine andesites and olivine-basalts of 

 the Kermadecs and Tofua in the Tonga group ; so that the absence of close 

 resemblance of the Savaiian and Hawaiian lavas can hardly be urged 

 against this hypothesis, even granting that a general similarity of type 

 could be expected from closely connected vents on the same line of 

 weakness — a condition which is very open to question. But there is cer- 

 tainly a Very close agreement between the White Island volcanics and 

 those of Ruapehu and Tongariro, as the following notes will show. 



In the hand-specimen all the samples are of dark-grey colour, and vary 

 from compact to scoriaceous. Small phenocrysts of feldspar and of the 

 pyroxenes are visible to the eye. The specific gravity varies from 2-60 in 

 the more scoriaceous varieties to 2-66 in the more compact. This corre- 

 sponds with MacLeod's specific-gravity determination of 2*65. 



Two specimens are from sea-level on the south coast. The first (No. 379 

 of the collection), when examined under the microscope, shows ordinary 

 andesitic characters. Phenocrysts of feldspar, hypersthene, augite, and 

 very occasional olivine are distributed in a groundmass of the pilotaxitic 

 type. The feldspar is a medium labradorite as determined by the statis- 

 tical method of Michel Levy, zonal structure being extremely common, 

 and inclusions or signs of alteration rare ; hypersthene crystals are ex- 

 tremely numerous, exhibiting straight extinction and moderately low polar- 

 ization colours ; augite occurs in well-defined crystals, frequently showing 

 twinning, and with high extinction-angles. The olivine appears as small 

 crystals, clear, with brilliant polarization colours, and occasionally, where 

 it was possible to obtain it, yielding the characteristic interference figure 

 of a mineral with an axial angle nearly equal to 90°. Magnetite occurs 

 freely as grains. The F.M. minerals are distributed throughout the slide, 

 but occasionally occur in nests composed of only one mineral, or of two, 

 or sometimes of all three minerals together. The groundmass consists' of 

 a felted mass of feldspar laths and microlites of either augite or hyper- 

 sthene ; some of these exhibit a greenish shade, and suggest the former 

 mineral. 



Another specimen from the south coast, at sea-level, is of the glassy type 

 of groundmass with a few feldspar microlites, and with a much smaller 

 proportion of hypersthene and augite. The feldspars, which are larger in 

 size than the average for the rocks, contain numerous inclusions of the 

 glassy groundmass. This shows numerous well-defined perlitic cracks, and 

 is rendered almost opaque with a dust of some dark-coloured mineral, 

 probably magnetite. 



Very similar to this one is a specimen from the base of the cliff on the 

 south side of the crater. Another sample, from the landslip at head of the 

 crater, is like the first rock described (No. 379) ; but occurrences inter- 

 mediate between the pure glassy type and the pilotaxitic type come from 

 flows at sea-level in Crater Bay and from the north coast. 



A flow from sea-level on the west coast of the island shows a divergence 

 from the usual characters of these andesites. It contains, in addition to 

 the usual plagioclase, a certain amount of sanidine, and the groundmass 

 is composed of untwinned feldspars of rectangular shape and augite grains 



