New. Hebrides, of 1913-1 k 537 



estimate of relative proportions in the most nearly holocrystalline 



varieties is — 



Phenocrysts, 8 per cent. 



Labradorite (Ab 3 A117) ..... 5 



Augite 2 



Olivine . . . . . . . . 1 



Groundmass, 92 per cent. 



Labradorite (Abi Ani) ..... 30 



Augite . . . . . . . t 45 



Magnetite ......... 5 



Glass + cryptocrystalline material ... 12 



"Augite is decidedly more abundant in slide Bd [from the west 

 end of the lava sheet east of Craig Cove] than in others of the 

 same rock. 



" Slide D is an olivine-basalt with much more abundant phenocrysts 

 than the above. Olivine and augite occur in large crystals only 

 slightly inferior to the felspar in quantity. Phenocrysts slightly < 

 or = groundmass. 



"The rocks are certainly basalts (olivine-basalts) in the modern 

 acceptance of this term. The distinction between basalts and 

 andesites is now made to rest more on the relative proportions of 

 light (felsic) to dark (mafic) constituents rather than on the nature 

 of the plagioclase or the presence or absence of olivine. The 

 andesites have a decided predominance of felspars over ferromagnesian 

 minerals, and the felspars are usually, though not necessarily, 

 andesine or acid labradorite. The basalts are characterized by 

 approximate equality of felsic and mafic constituents, and generally 

 by basic plagioclase and presence of olivine. Thus one may have 

 olivine-andesites with phenocrysts of basic plagioclase (a type very 

 common in the circum-Pacific volcanoes), and basalts devoid of olivine 

 and containing only moderately basic felspars." 



The specimens collected by Professor Davis from Dip Point 

 (Marshall, 1915, p. 391) and those brought by Mr. Prater from 

 Harbour Crater agree in character with those from the lava streams 

 discharged from the fissures at Po-luk. They are all basalts of the 

 same type. Hence the explosive or non-explosive nature of the 

 eruption at Ambrym must depend, not upon the nature of the rock, 

 but upon the conditions of its eruption, and probably on the character 

 of the surface into which it was intruded. 



Professor Lacroix refers to the specimens he examined as too poor 

 in olivine to be true basalts, although they contained 5 per cent of 

 olivine. This amount is above that in most of Mr. Frater's 

 specimens, in which Mr. Tyrrell determined the percentage of 

 olivine as only one. Professor Lacroix identified the specimens he 

 investigated as andose. The identification is of course correct, but 

 the name is inappropriate, since, according to Professor Iddings 

 (e.g. 1913, ii, p. 615), andose is a basalt and not an andesite, the 

 natural affinity suggested by the name. Probably the older Ambrym 

 lavas associated with the more explosive eruptions of the prehistoric 

 volcanic eruptions on the island may be andesites ; but they are not 

 represented in Mr. Frater's collection, which, as Mr. Tyrrell's note 

 shows, contains nothing that can be called an andesite. 



