54 PROCEEDINGS OF THE WASHINGTON MEETING 



Several determinations for specific gravity were made and the results were 

 found to vary considerably. Examples are 1.00, 1.25, and 1.36. In making 

 these determinations it was desired to obtain the density of the body — that is, 

 glass sponge with included air, rather than that of the glass composing the 

 body. 



Chemical Constituents 



In order to determine whether these bodies represent a special differentiated 

 phase of the magma, with greater silica content and higher viscosity or other 

 particular properties, a partial analysis was made. The writer is indebted to 

 L. J. Youngs, of State College, Pennsylvania, for the following figures : 



Per cent 



Si0 2 49.94 



A1 2 3 13.52 



FeO (total iron calculated as ferrous iron) 10.79 



Ti0 2 1 . 26 



CaO 9.74 



MgO 11.80 



Total 97.05 



There was a very slight, gain on ignition. The alkalies were not determined, 

 owing to scarcity of material on hand, but by difference they would correspond 

 closely with amounts found by other analysts in rocks from this region. The 

 rock is a basalt in composition, and under the microscope the crushed material 

 is found to consist of 70 to 90 per cent of brownish glass, the proportions vary- 

 ing in different bodies, with small phenocrysts of feldspar and pyroxene and 

 possibly some olivine. 



A comparison of the analysis with various analyses of Kilauean rocks found 

 in Washington's tables 6 and made by Silvestri, Phillips, and Lyons shows that 

 the magnesia is very much higher and, as a rule, the lime is considerably 

 higher than for most of these rocks. It compares better with an analysis of 

 Pele's hair made by A. H. Phillips and is similar in most respects to an analy- 

 sis by Merwin of the lava dipped from the crater Halemaumau. 7 



Although the figures obtained by Mr. Youngs were carefully checked and 

 there seems to be an established high percentage of magnesia, it is not safe to 

 draw conclusions, without further analyses, as to whether a high magnesia 

 content is always characteristic of "Pele's tears." If it be a special feature, 

 then may not their composition have some bearing on the question of the tem- 

 perature at which this lava becomes liquid and possibly on the origin of these 

 particular bodies? This point is more interesting because of the proportionally 

 high percentage of lime and magnesia in the austral ites and the high tempera- 

 ture at which they dissolve and pass into the liquid condition. The calcium 

 and magnesium bearing minerals, such as the olivines, pyroxenes, and more 

 calcic feldspars, are always considered among the high-temperature minerals. 



8 Chemical analyses of igneous rocks. Professional Paper 14, U. S. Geol. Survey. 

 7 Day and Shepard : Water and volcanic activity. Bull. Geol. Soc. Am., vol. 24, 1913, 

 p. 586. 



