166 Jaggar — YolcanoJogic Investigations at Kilaiiea. 



is furnished by the cracking, streaming and foundering pro- 

 cess i^o^. 12d) whereby great quantities of porous hardened lava 

 crust carry air down into the gas-charged melt and probably 

 discharge it rather slowly, owing to the sealing of the pores 

 by chilled glass as the slabs first sink in the liquid just below 

 the surface. This liquid is at a temperature so low (see experi- 

 ments hereafter) as to be solidified by such contact, and has no 

 power to melt up the deeply chilled crusts which are several 

 inches thick. A fourth mechauism which brings oxygen and 

 volcanic gases into contact beneath the surface of the lava col- 

 umn is furnished by all those cavings-in and crackings of the 

 older rock whereby avalanches (fig. 4(2) or single blocks are 

 precipitated into the liquid lava, or where the liquid lava by 

 percolation through newly-opened crevasses, or by " stoping," 

 gains access to broken surfaces of fragments filled with air. 

 A fifth process, not at all obvious but possibly very effective, 

 is the indraught of air from the wall tlirough the pores of the 

 lava column, creating a blast furnace and incessantly com- 

 pensating a tendency to vacuum created by chemical reactions,* 

 or by convectional gas pumping, within that column. 



It is interesting to note that Wm. Lowthian Green in his 

 " Yestiges of the Molten Globe" held that the fountainiug of 

 the lava at places of descent was due to air being carried down 

 with the lava and expanded. He lays great stress on the quan- 

 tities of air carried down in the lava and even insists that vesi- 

 culation is due to air. Expansion of air would not account for 

 the flames observed, and the temperature is not high enough 

 for dissociation of water vapor, which Green insisted on. 

 ]>J either will quiet foundering of crusts, however, nor a lower- 

 ing of the general level by escape of gas, as suggested by Fer- 

 ret and Daly, account for the violent downsucking at the 

 fountains and grottoes. This action is frequently sudden, or 

 graduated from slow to fast in a very short distance. Continu- 

 ous or spasmodic disturbance of equilibrium of the lava by 

 oxidation of gases with devesiculation would account for the 

 phenomena observed. 



The chief gases collected at Kilauea in 1912 from a flaming 

 cone on the floor by Day and Shepherd had the following 

 approximate average compositionf by volume for 1000 liters of 

 gas which was pumped : 



Water 4fo 



Sulphur dioxide. 50 



Carbon dioxide 25 



"Dt. Shepherd informs me that in the reaction 2H2 + Oo = 2H2O, the vol- 

 ume shrinkage is one- third even at high temjjeratnres. 



f Kindly estimated for the writer by Dr. A. L. Day. The SOo is donbtfnl, 

 much of it being probably free S. Assuming derivation from air, this partly 

 accounts for the excess of combined for the N joresent. 



