272 8. Powers — Notes on Hawaiian Petrology. 



limestone may be necessary to canse snck differentiation 

 in a large amount of basalt, there are grave difficulties 

 connected with the application of the theory of lime- 

 stone control to islands which, so far as known, are built 

 almost wholly of basalt. It is true that limestone occurs 

 as a thin veneer on the margins of Oahu; that a very 

 small amount of elevated limestone occurs around Kauai 

 and Xiihau, and that a single elevated reef has been 

 traced a short distance on southwestern Molokai. Mod- 

 erately wide coral reefs surround portions of Xiihau, 

 Kauai, and Oahu, but the reefs around the other islands 

 are of very limited extent. The tuff cones Ulupau 

 Koko Crater, Makalapu, Punchbowl, and Diamond Head, 

 on Oahu, contain fragments of coral, but only from the 

 immediately underlying reef, and no cones on Molokai, 

 Maui, Hawaii, or Kauai are known to contain such frag- 

 ments. The greatest difficulty with the theory is the fact 

 that no limestone has been found interstratified with the 

 basalt flows in any canyon on the Islands, although 

 erosion has exposed many sections 2,000 to 3,000 feet be- 

 low the former summit crater. 



In order to have limestone available for fluxing a 

 basaltic magma, it must occur at a depth of several 

 hundred feet, for lava is injected as dikes and through 

 fissures to the surface with great rapidity. Limestone at 

 the required depth would apparently have to represent a 

 deep-sea deposit of considerable thickness rather than a 

 coral-reef deposit unless this portion of the Pacific gradu- 

 ally subsided as the volcanoes were built up. 23 That such 

 deep-sea deposits exist in the lower portions of the huge 

 volcanic piles is subject to discussion. 



Nephelite basalt. — Rocks containing nephelite, usually 

 as a very subordinate constituent, have been identified on 



23 H. A. Pilsbry has accounted for the present distribution of the land 

 shells of the family Achatinellidae over the Hawaiian Islands by submer- 

 gence during the Late Pliocene or Pleistocene, progressively separating the 

 various islands. He maintains that the differentiation of the modern fam- 

 ilies of land shells took place largely in the Mesozoic and that "the hypoth- 

 esis that the Hawaiian volcanoes rise from a pre-existing mid-Pacific ridge, 

 now lost by subsidence, gives room in time and space for the derivation of the 

 peculiar fauna" (p. xlvii) (Manual of Conchology, Second Series, vol. 22, 

 pp. xlii-xlvii, 1912-4; also quotation and map by W. A. Bryan, Natural His- 

 tory of Hawaii, Honolulu, 1915, pp. 121, 291). P. Marshall, B. Koto, and 

 others, however, have shown that andesites are the type of volcanic rock 

 associated with continents, and Marshall maps the northern limit of the for- 

 mer continent of Oceania on this basis as passing not far north of Fiji 

 (Oceania, Handbuch der regionalen Geologie, Heidelberg, 1912, Bd. 7, Heft. 

 5, fig. 3, p. 5). 



