268 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



which has so h)!!*; filled this throat to varvin*^ depths from a few 

 feet to as much as 700 feet or more. 



To those who have thought of the volcano crater as a downward ex- 

 tendin*; i)ii)e or conduit throu^li whidi the lavas for centuries have 

 been quietly poured out upon the surrounding terrane the appear- 

 ance of the enlarged and deepened opening must have been an 

 unexpected revelation. It api)eared quite dry — no freshly cooled 

 volcanic glass (obsidian) or partially crystallized lava was visible 

 below, nor was any found among the fragments thrown out (pi. 3, 

 fig. 1) by the explosions. The lava fragments were from older layers 

 than any before exposed at Kilauea, completely crystallized, nearly 

 free from bubbles, rather fine grained though rich in olivine and feld- 

 spar phenocrysts like the lavas from its great neighbor, Mauna 

 Loa, but in phj^sical appearance totally unlike the Kilauea flows 

 which have almost come to be regarded among geologists as a fixed 

 type. 



At one side of the great bowd, about GOO feet above the bottom, 

 was an area some 500 feet in its transverse dimension and more than 

 100 feet thick, which showed here and there a trace of red at night. 

 The surface also had the appearance of fresh Aa lava, which occa- 

 sionally flaked off in considerable areas, revealing a bright red patch 

 beneath. Above it on the rim a hot air-current was continually 

 depositing fine flakes of freshly-oxidized iron-bearing scale. This 

 may have been one of the feeders of the lava lake. It was cer- 

 tainly the hottest spot left exposed in the empty basin. Another 

 smaller area in one corner of the bottom was distinguished by a half 

 dozen roaring gas outlets whose throats glowed red at night. This 

 may have been a smaller feeder. When the lava began to return to 

 the pit in July it spouted out from a point high up on the talus 

 pile on the opposite side of the basin in a fountain 175 feet high. 

 This must have been from a third feeder. No others have so far 

 been discovered. 



Instead of the boiling-flask picture, therefore, we should think 

 rather of a central collecting tube, w'ith many more or less wide- 

 spreading branches below, leading to local chambers in which crys- 

 tallization is proceeding under different conditions of temperature 

 and pressure. That a number of such branches exist is only dis- 

 covered on a rare occasion like the present summer wdien violent 

 explosions have cleaned out the lava basin and enlarged it to many 

 times its former volume. Then it is seen that the openings leading 

 into and out from it are on the side walls, while the bottom is to all 

 appearance solid, old, and (relatively) cold. Even the solid ejecta 

 which were thi-own out during this period of violenfactivity include 

 no obsidian masses and no lava as recent as that represented in the 

 lava lake and in the flows of recent years. 



