Jaggar — Yolcanologic Investigations at Kilauca. 203 



ing over a solid or semi-solid obstruction. And this was just 

 at the level where the irregularities of the margin of the lake 

 bottom of the previous higher stand of the lava might be 

 expected to appear. Incidentally the cavern was at a location 

 of large and continuous border grotto building for many 

 previous months. 



During the progress of such cascading the lake may main- 

 tain its level, or even rise, showing that the cascade is not an 

 outlet. It merely reveals during excessive shallowness, con 

 vectional mechanism commonly concealed by greater depth in 

 the lake. That the violent downrush during cascading, how- 

 ever, is swifter than at other tinies, is shown by the fact that 

 when such a cascade is in action, it is apt to satisfy completely, 

 or nearly so, the fountaining requirements of the lake mechan- 

 ism, for almost all other fountains cease playing while the cas- 

 cade is falling. This might be expected, as the rest of the 

 lake at such times is very stagnant and crust foundering is 

 largely limited to the one place where crusts and air are being 

 sucked down in prodigious quantities and the heat and flaming 

 are excessive. Probably excess of downward convection over 

 inflow or some equivalent mechanism in the relation of bench 

 magma construction to lake magma eflervescence, brings about 

 the cascade phase of downflovr. If, however, it is maintained 

 by aa extreme phase of oxidation of gases, which keeps the 

 level of lava in the sinkhole 20 to 50 feet (6 to 15 meters) 

 low^r than the level of the lake, then we have to make careful 

 inquiry as to how this could be brought about. 



Condemnation by de-vesiculation. 



In the above discussion of oxidation fountaining, it was 

 pointed out that two opposed processes w^ould operate in oppo- 

 site senses, when foundering crusts along with convection 

 downflow mixed air with explosive gases, and so induced oxida- 

 tion, excessive heat, consequent gas expansion, and consequent 

 locahzed upward gas convection. Mention has been made in 

 addition of condensation resulting from the explosion of un- 

 stable mixtures of oxygen with hydrogen, carbon monoxide 

 and sulphur vapor. Wliether there would be such condensa- 

 tion at the high temperatures prevailing in molten lava is a 

 question for the physical chemist. We know that two atoms 

 of hydrogen uniting with one of oxygen produces a molecule 

 of water which in its liquid state occupies greatly diminished 

 space in contrast to its gaseous progenitors. In the case of 

 water vapor at 850° C, however, condensation would be less, 

 and the same is true of the transformation of sulphur with 

 fren to S0„ and of CO with oxyj^ren to CO,. If in the 



oxy 



'2 ciii^ v/x v^v^ »>^i..x y^vjj^ 



heated condition there is less change in the vapor tension 

 resulting when these reactions take place, it is evident 



