48 PROCEEDINGS OP THE WASHINGTON MEETING 



DIFFUSION IN SILICATE MELTS 

 BY N. L. BOWEN 1 



(Abstract) 



This paper gave a brief description of some determinations of the rates of 

 diffusion of molten rock-forming silicates and a discussion of the significance 

 of the results in petrologic problems. 



Presented in abstract extemporaneously. 



Discussion 



Prof. J. 1'. Iddikgs remarked that Mr. Bowen's investigations were made on 

 stationary liquids, but it is not to be supposed that large bodies of magma 

 having different temperatures in different parts will exist without convection 

 currents for any appreciable length of time, and his remark that convection 

 currents may play a prominent role in differentiation processes is very much 

 to the point. The effect of gaseous constituents in molten magmas, or of other 

 components which reduce the viscosity of such liquids, must also be taken 

 into account. 



PETROGRAPHY OF THE PACIFIC ISLANDS 

 BY K. A. DALY 



(Abstract) 



The total number of named islands in the open Pacific is about 3,000. Of 

 these, only 22 are reported to show outcrops of quartzose rocks. Five other 

 islands are reported to have outcrops of crystalline schists, serpentine, or de- 

 formed limestone. 



Excluding New Guinea and New Zealand with their immediate satellitic 

 islands, Oceania has 345 islands which have been definitely described as wholly 

 or largely volcanic in origin. Probably the whole number showing volcanic 

 rocks above sealevel is at least twice as great. Among these islands only 15^ 

 have yet afforded any petrographic data, and not one has been examined with 

 desirable thoroughness. To illustrate the scrappiness of our information as 

 well as certain problems regarding the origin of Pacific lavas, the igneous-rock 

 types of each island, so far as recorded, have been tabulated. 



The compilation illustrates the advisability of a systematic exploration of 

 all the smaller islands of the Pacific. The task is quite feasible, since the 

 total land area involved is less than 75,000 square miles. If Hawaii, Viti 

 Levu, Vanua Levu, New Pomerania, New Mecklenburg, and New Caledonia be 

 excluded, the total land area to be covered would be only 50,000 square miles. 

 In fact, with relatively small effort and cost, the natural history (including 

 the petrography and geology) of all the land areas within one-eighth of the 

 earth's surface could be investigated by a single organization. The Geological 

 Society may well use its influence in developing a plan for such comprehensive 

 exploration in the Pacific, under private American auspices. 



Presented in abstract extemporaneously. 



1 Introduced by C. N. Fenner. 



