Halemaumau Lava Lake. 



347 



again, being now entirely under the influence of the surface 

 current. 



The reason why the normal lava lake will have a greater 

 length in some direction away from the conduit, and why the 



Fig. 5 : A (above) and B (below). 



Fig. 5. Comparative views. A. July 5, 1911. B. Sept. 9, 

 meters lower. Showing change of form through lesser activity. 



1911 : 25 



lava will circulate in the direction of its length, is now made 

 beautifully clear by the present hypothesis. From the diagram 

 it will be seen that the downflow at the east consists of the 

 coolest lava which, even if it does not consolidate to the walls 

 by reason of its rapid motion, is, at all event , quite lacking in 

 re-fusing power, while the large hot gas bubbles only rise at 

 intervals — the wall at this end remains, therefore, much after 



