258 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1925 



It is iinportanf to note fui'tlier that islands are fre(|uently present 

 in the hike which liave an upward and downward nio\ement of dif- 

 ferent character and period from tliat of the fluid lava; also that 

 when a fall of the fluid lava f)ccurs it often exposes what appears to 

 be solid material only a few feet below the previous liquid surface, 

 with no relation to the 700-foot level or to any deep-seated point. 

 Such observations compel the admission that the material contained 

 within the lava basin is considerably diversified in its physical state ;^ 

 some is very fluid and some nearly or quite rifjid. 



This observation led to a rou<;h lal)oratory test of some signifi- 

 cance. On those occasions when the lake overflows its banks, which 

 is a matter of daily occurrence during the rise of the lava column, 

 there is visible movement of the fans and rivulets of new lava over 

 the crater floor until they are so far cooled clown that hardly a 

 trace of red glow from them can be seen. The lowest temperature 

 at which movement can be detected is therefore of the order of 600°. 

 A fiagment of this same lava brought to the laboratory and re- 

 heated w'ill not flow under its own weight until a temperature in 

 the neighborhood of 1,300° has been reached. Such an observation 

 is significant of a change of composition and yet the only change 

 of composition i:)Ossible must be in the amount of volatile matter 

 (gas) in solution in it. Gases are given off in abundance during 

 cooling and their loss takes away from the lava its fluidity over a 

 considerable range of temperature. 



This observation taken with the other perhaps explains why the 

 physical state of the lava in the basin is thus discovered to be partly 

 rigid and parti}' fluid, partly porous and partly dense — its gas 

 content is variable ; and a further consequence is altogether probable, 

 namely, that there are differences of temperature throughout the 

 mass of material filling the crater greater even than the observed 

 differences at the surface. 



Moreover, during this period of continuous observation in 1912 

 it was uniformly noticed that an increase in the amount of gas dis- 

 charged at the surface of the lake was always accompanied by a rise 

 in surface temperature, and conversely during falling temperature 

 the gas fountains diminished in number. A photograph taken at 

 a time when the temperature was highest (July 3, 1912) offers a 

 distinct record of more than 1,100 fountains distributed almost uni- 

 formly over the surface of the lake (pi. 1, fig. 2) wdiile at other times 

 no fountains at all are to be seen for intervals of a minute or more, 

 and then only an occasional one appears. This observation imme- 

 diately led us to the conclusion that there is a causal relation be- 



"T. A. Jaggar. Seismometrlc Investigation of Hawaiian Lava Column. Bull. Seism. 

 Soc. Am. 10, 160, 1920. 



