P. A. Daly — Problems of the Pacific Islands. 163 



the huge field lying open for future research in this branch 

 of geology, I have prepared a table of the Pacific islands, 

 already recorded as volcanic ; and have entered therein the 

 names of the different rock types, so far as determined by com- 

 petent petrographers. It is planned to publish this long table 

 elsewhere. 



Origin of the lavas. — From twenty-one out of the twenty- 

 eight more important archipelagoes, common feldspar basalt 

 has been reported. Of the remaining seven, the Mariana group 

 is the only one from which any petrographic data seem to have 

 come into print; this archipelago is so little known that it can 

 not yet be regarded as free from basalt. In all, eighty-one 

 islands are said to contain feldspar basalt or its chemical equiv- 

 alent, diabase or gabbro. 



Pyroxene andesite is known to accompany the basalt in nine- 

 teen islands and is found in thirty-three others. The two 

 types are not only present in the form of closely associated 

 lava flows ; they are also connected by transitional types, called 

 basaltic andesite, andesitic basalt, or olivine andesite. In both 

 respects the dominant Pacific-island lavas are like the dominant 

 lavas in the other ocean basins and on each of the continental 

 plateaus. Their geographical and mineralogical relations sug- 

 gest that basalt and pyroxene andesite are connected in origin. 

 What is that connection? 



According to the only view which seems well supported by 

 physical, chemical, and geological facts, the andesite is best 

 regarded as a direct derivative of common basalt. Under cer- 

 tain conditions, some of the heavier constituents settle out of a 

 column of slowly cooling liquid basalt, leaving a less dense 

 liquid of andesitic composition at the top of the column. The 

 ordinary cylindrical vent of a volcano has the form of such a 

 column ; if the temperature is held long enough within the 

 right limits, andesitic liquid is developed in the upper part of 

 the vent, whence intermittent outflows of the new type of 

 lava are easily possible. This hypothesis is important, since it 

 refers to that kind of lava which, next to basalt, is probably 

 the commonest on the earth. If it be correct, the overwhelm- 

 ing preponderance of primitively basaltic material in terrestrial 

 eruptions is all the more evident. The Pacific islands are 

 already known to furnish an excellent field for further testing 

 of this theory. 



The archipelagoes also illustrate the close genetic connec- 

 tion between common basalt and many types of lavas which 

 are often described as belonging to the alkaline series. What 

 that connection may be is a difficult problem on which I can- 

 not now enter. Suffice it to say that Hawaii, Samoa, Tahiti, 

 Juan Fernandez, the Hervey islands, and the Fijis have given 

 valuable information on the subject, though in no case is either 



