340 



F. A. Perret— Circulatory System in the 



through the less resisting " bright lines," forming tiny foun- 

 tains, and the rest, accumulating under the crust, issue as a 

 continuous gas flow through innumerable holes and rents, burn- 

 ing in the air with the steady, pointed, blow-pipe flames 

 described in the paper on lava fountains. 



On arriving at the eastern portion of the lake the lava has, 

 therefore, suffered three principal modifications, all of which 

 contribute to one grand result which, as we shall see, is the 

 key to the problem of the circulation — it has lost its gases, it 

 has lost heat and has become, in part, solidified. The net 



Fig. 1, 



Fig. 1. Night view. Bright lines formed by the pulling apart of the 

 crust. Western half of lake. From the North. 



result of all this is that the surface layers are denser and 

 heavier than the rest of the lake material and therefore tend 

 to sink. This is most evident in the crust itself which is a 

 solid crystalline rock, pieces of which sink readily in the liquid 

 as is forcibly shown during the frequent upward oscillations of 

 the lake. It might reasonably have been supposed that a 

 rising lava column would raise its floating crust but this it does 

 not do. The crust breaks by uneven pressure from below and 

 the lava rises through the cracks, up-tilting the broken pieces 

 which readily sink edgewise in the liquid. The rising lava 

 does not raise the crust but breaks it up, rises through it, and 

 sinks it. 



