Marshall. — Geology of Tahiti. 375 



from magmatic differentiation is a question for the future. . . . The 

 constant association of the main types in the islands visited truly supports 

 this view.'"* 



Interest largely centres in these oceanic islands in connection with the 

 structure and origin of the Pacific basin. Many attempts have been made 

 within recent years to establish an association between coastal structure 

 and the products of volcanic activity, notably between the coast lands of 

 the Pacific types of Suess and the occurrence of andesitic rocks. This idea 

 has been supported by Harker and Prior, amongst others, while Gregory 

 and others have dissented from it. Thus Lacrorx points out that the ande- 

 sitic type is strictly circum-Pacific. This statement holds in the present 

 state of our knowledge, as pointed out by Marshall, if, as seems reasonable. 

 Tonga, Fiji, New Hebrides, Santa Cruz, Solomon, Caroline, and Marianne 

 Islands are considered as circum-Pacific in their situation, a suggestion 

 which is strongly supported by ocean soundings. 



On the other hand, those islands which are situated in the central areas 

 of the Pacific Ocean are, so far as known, constituted of the alkaline and 

 basic facies which are referred to the Atlantic region. It is, however, possible 

 that the Marquesas (biotite trachyte, Lacroix) and Easter Island (andesites, 

 Helsch) are exceptions. In order to explain this contrast between the ande- 

 sitic border and the alkaline basic central area, the suggestion of Supan 

 that andesitic rocks are the product of eruption where rock-folding is in 

 progress, and alkaline basic rocks where radial fractures have been formed, 

 demands attention. There is, however, so little knowledge of the actual 

 structure of the basement on which these island groups are situated that 

 any stated conclusions are more likely to be guesses or the expression of 

 the personal bias of an author than scientific deductions based on. reliable 

 premises. 



Other questions arise. Is the basin of the Pacific Ocean the scar left 

 by the moon when it separated from the earth, an infallen area, or a de- 

 pression of an isostatic nature due to the high, specific gravity of the materia] 

 of the earth's crust in that region ? The study of the petrology of these 

 island groups ought to produce some evidence on this point confirmatory 

 or otherwise of such pendulum experiments that have been. made. 



I do not know whether the first theory should or should not require any 

 particular nature of the rocks forming the floor and the material immedi- 

 ately beneath the floor of the ocean-basin. If there is, any one would perhaps 

 expect the rock to be of an ultra-basic nature. If the area is infallen, it 

 would be reasonable to expect to find occasional fragments of sediments 

 included in the tuft's or lava-flows, as at Mount Ruapehu, New Zealand, 

 as well as at Auckland and Dunedin. Search has, however, failed to reveal 

 any such fragments in the islands of the Society and Cook Groups that 

 I have visited. It would thus appear that sedimentary rocks form a less 

 important part of the earth's crust in the neighbourhood than in the areas 

 of the great volcanic districts of New Zealand. 



On the other hand, there is in the volcanoes of these islands a great 

 predominance of basic types of rock. In the basic rocks feldspar is relatively 

 unimportant, and is in almost all cases confined to the groundmass. In 

 the phonolites, which are. however, found in much smaller quantity than 



* Marshall, P., " Handbuch der Regionalen Geologie," band vii, abt.2 ; also Rep. 

 Aust. Ass. Sci., vol. 13, 1912, p. 201. 



