MAGMATIC DIFFERENTIATION IN HAWAII 299 



According to the Mode classification, it is an andesitic basalt, 

 transitional in type between olivine basalt and augite andesite. 

 Its specific gravity was determined on a specimen which had been 

 coarsely powdered to avoid an error due to the porosity of the 

 rock. For comparison the calculated average for the basalt of 

 Hawaii is given in column 2 . 



Trachydolerite of summit flows, Mauna Kea. — Brigham and 

 others long ago noted the occurrence of "clinkstone" at the top 

 of Mauna Kea, and the present writer had opportunity to make 

 some study of this rock in place during the 1909 reconnaissance. 

 Near the 13,000-foot contour, he found several flows of lava of 

 much lighter color than the staple olivine basalts of Hawaii, or 

 than the abundant andesitic basalt just described. These flows 

 all seem to be short, generally less than one kilometer in length. 

 Their terminal scarps have been little affected by frost or other 

 weathering agents, and the steepness of these scarps indicates a 

 notable degree of viscosity during the outflow. In some cases 

 these flows could be seen to have emanated from the fissures in 

 the summit cinder-cones. Though the pyroclastic material of 

 the cones is generally altered (to deep brown and red tints), it 

 appears to be chemically identical with that composing the always 

 fresh, light-colored flows. 



The "clinkstone" habit is due largely to a noteworthy lack of 

 vesicles in the lava. Though some large gas-pores always occur 

 in the thin surface shell of each flow, its interior is often nearly 

 or quite free from even small pores. This homogeneity of the 

 rock is, doubtless, chiefly responsible for the extremely sonorous, 

 metallic sound given out when the lava is broken by the hammer. 



For special examination, typical specimens were taken from 

 a flow which issued from the eastern flank of the cinder-cone 

 named "Poliahu" on the government map, at a point about 

 350 meters north of the summit pond. The description of the 

 lava may be based on one specimen, which has been chemically 

 analyzed. 



The rock is of a fairly light, slate-gray color, is non-porous, very 

 dense, but holocrystalline. A few thick tables of plagioclase, from 

 1 mm. to 2 mm. in length, represent the only constituent determi- 



