202 Jaggar — Volcanologic Investigations at Kilauea. 



during risino-, m the form of a Y with arms curved outward, 

 the stem being the current from the source. This is what 

 happened in the summer of 1916 when a simple oval lake 

 developed powerful streaming outward from tlie west side, 

 then shoals appeared and peninsulas were built out (fig. 5) on 

 both sides of the current, until the Y or T form was produced 

 (tig. 6a). 



Sinkhole Cascades. 



The extreme case of the grotto fountain is the sinkhole cas- 

 cade such as was described on pp. 172, 173 above. Such cascades 

 from the lake into holes at its border, or in its bottom, have 

 repeatedly appeared when general subsidence began. Com- 

 monlv the place of cascading is a border pot which may have 

 been a place of rising lava pouring out into the lake during a 

 previous rising spell. When sinking begins the magma in the 

 smaller tube sinks lower than the lava in the lake with the 

 result that a torrential fall is precipitated from tfie lake into a 

 glowing void, intense effervescence or fountaining is in progress 

 in the depths of the pot which receives the fall, great sheets of 

 crust from the lake are drawn over the rim, and break up in 

 the abyss, and a column of flame and fume rises above with 

 great heat and much sulphurous acid gas (fig. ^b). All the 

 phenomena of such a cascade are those of a border grotto 

 exaggerated (fig. llci"). 



For a long time in watching such cascades the writer was in 

 doubt whether the lake lava actually poured over a ledge or 

 whether the appearance of such an q^^q were not merely the 

 more rapid fall of surface layers over the slower liquid beneath 

 where the magma in the border fissure, open on the lake basin 

 side, sank as a froth by loss of gas faster than the magma of 

 the larger lake body. 1 am now convinced that commonly 

 there is an actual lip of bench magma which the lava of the 

 lake pours over, and this lip is usually the margin of the lake 

 bottom, revealed by reason of the sinking of the liquid lake 

 within its basin in the bench magma. 



There has recently (February 21-28, 1917) been a remark- 

 able case of such cascade action after 46 feet (14 meters) of 

 subsidence of the lava lake — just sufficient to uncover much of 

 the bottom and leave the lake a very shallow body coursing 

 like a river. At the northeast margin under the cliff of bench 

 lava there was revealed a cavern in the w^all into which the 

 lake lava cascaded for seven days and longer. The fall was 

 shaped in plan like an obtuse Y w^itli the point toward the 

 lake ; the cascade poured over both arms of the Y, while in 

 one place an outcropping of ledge rose through the fall and 

 divided it. This clearly indicated that the liquid was cascad- 



