354 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1960 



new land beyond the former coastline. Between February 6 and Feb- 

 ruary 20 the main vent area continued to emit gas, spatter, and pumice, 

 building a large cone to a height of 420 feet above sea level. 



The first lavas extruded in the flank eruption were very similar 

 chemically and mineralogically to the lavas erupted in 1955, and it is 

 highly probable that they represent magma stored in a relatively near 

 surface reservoir after the activity in 1955. Both lavas are poor in 

 olivince but contain abundant phenocrysts of plagioclase feldspar and 

 pyroxene. Their degree of differentiation is also shown by the rela- 

 tively high silica content of about 50 percent. Near the end of Jan- 

 uary, after 2 weeks of steady eruption, the nature of the lava gradually 

 changed. The temperature increased to 1,100° C, compared to 1,050° 

 to 1,060° C. for the earlier material, the viscosity decreased, and 

 olivine phenocrysts became abundant. Apparently the magma which 

 was forcing its way through the rift zone from the summit was only 

 then reaching the flank eruption area. During the last week of strong 

 lava output in February, the temperature increased even more — maxi- 

 mum measured was 1,130° C. — and the lava was nearly identical to 

 that of Kilauea Iki. 



Shortly after the outbreak of the flank eruption tilt measurements 

 revealed a very rapid settling of the summit area (inset B, fig. 2), and 

 by the end of January a strong swarm of shallow earthquakes was 

 being recorded at the summit stations. As more and more lava moved 

 from the summit area through the rift zone toward the flank eruption, 

 the settling became even more pronounced and the earthquakes more 

 frequent. Strong, frequent earthquakes from the vicinity of 450-feet- 

 deep Halemaumau, together with the issuance of steam from new areas 

 around its floor, led Survey scientists to believe that the rapid with- 

 drawal of lava from beneath the caldera might result in some form of 

 surface collapse. By daybreak on February 7 the collapse had begun, 

 and within a few hours the former flat floor of Halemaumau was a 

 saucer-shaped depression 150 feet beloAv its original level. Then just 

 before noon and within a period of 10 minutes a small area, 1,000 feet 

 in diameter in the center of the larger collapse feature, dropped an 

 additional 200 feet, and a small volume of highly viscous lava oozed 

 into the newly formed hole. Through the afternoon the subsidence 

 gradually ceased and the oozing of lava abated. Two days later 

 another smaller collapse occurred in the floor, but was not accom- 

 panied by the extrusion of lava. All the collapse appears to be di- 

 rectly attributable to the withdrawal of a portion of the still fluid 

 core in the 300-foot lava lake which filled Halemaumau in 1952. Some 

 lava, however, was forced outward into the innermost collapse pit by 

 the lithostatic pressure being applied around the periphery of the 

 main collapse area. 



