326 R. A. DALY PETROGRAPHY OF THE PACIFIC ISLANDS 



and anthropology were thus fully worked out, the results for general 

 science would be incalculably beneficial. 



This paper is intended to emphasize the rich field for research to be 

 found in the petrography of the Pacific islands. It is possible to sum- 

 marize in a few pages nearly all that has been done so far in this field. 

 The facts are most readily assembled in tabular form, and the three fol- 

 lowing tables tell the story of published petrographic results up to Jan- 

 uary 1, 1916. The tables primarily refer to Oceanica, but some other 

 islands are included. New Zealand, New Guinea, and their satellites, 

 as well as the islands close to continents, are not included. The tables 

 are doubtless not quite complete, but they serve to stimulate the active 

 petrologist. In spite of the very small amount of systematic work yet 

 accomplished, the published facts can not fail to suggest certain questions 

 regarding the origin of the Pacific lavas and other island rocks. Some 

 of the problems will be briefly considered, particularly those dealing with 

 the volcanic islands. 



"Continental" Eocks 



Table I lists the thirty-three or more islands which are now known, 

 or have been suspected, to contain quartzose rocks, basic schists, serpen- 

 tine, or deformed limestones. These are all rocks that abound in each 

 of the continents and, for convenience, they may be called "continental 

 rocks." Ellis's statement that he found quartz-feldspar rock on Borabora 

 of the Society Group and Commander Thomson's reported discovery of 

 granite and slate on Easter Island are both doubtful and need testing in 

 the field. With those exceptions, all the Oceanican continental rocks 

 so far known to the writer in published record lie west of a straight line 

 joining the eastern Fijis with the Mariana Group. In principle this dis- 

 tribution is, of course, well known and has long been explained in con- 

 nection with the theory of Tertiary fragmentation of the Australia- 

 Melanesia continent. 



Volcanic Islands 



The much longer Table II summarizes present-day knowledge of those 

 open-ocean islands which are wholly or largely composed of volcanic rocks. 

 To date 370 islands have been definitely so described. The whole number 

 of them probably well exceeds 600. Of these only about 180 have fur- 

 nished any information regarding the nature of their constituent rocks. 

 Not a single island has yet been studied and mapped in the detail suited 

 to the needs of petrology. 



In Table II there are 82 names of rock species. Most of these are 

 readily grouped into four classes, respectively representing basaltic, ande- 



