F. A. Ferret — Floating Islands of Ilalemaumau. 275 



"We may conclude, therefore, that the majority of the larger 

 and more important of these islands are formed in this simple 

 manner, analogous, it will be seen, to the making of icebergs 

 by the detachment of masses from a glacier on its entering the 

 sea and, similarly, constituting one phase of a complete and 

 recurrent cycle of events. The water of the sea, evaporated, 

 falls as snow in the uplands, is consolidated to glacier ice 

 which descends and is re-melted in the sea. The liquid lava 

 rises in the conduit, is consolidated to the walls of the pit ; the 

 liquid subsides, the masses of rock break off and fall into the 

 lake, floating about until swallowed up and re-melted in the 

 seething cauldron. 



The analogy fails, however, in regard to the flotation of the 

 mass for, if the specific gravity of water in the solid state is 

 less than in the liquid, the same cannot be said of lava. And 

 this is especially true in the case of the lake material whose 

 gas-vesiculated upper layers, as has been shown in the paper 

 on Lava Fountains, offer but little resistance to the sinking of 

 comparatively small pieces of the consolidated rock. It may 

 well be considered strange that, where smaller blocks sink, a 

 large mass of the same material, heavier than the liquid, should 

 be floated upon it, and the first questions of the visiting 

 tourist are generally " why does not the island sink and why 

 does it not melt V The answer is that it does sink and it does 

 melt, but not all at once — it may be a matter of months rather 

 than of moments. A number of causes may combine to bring 

 about the initial flotation, and some of these will be effective 

 in refloating the mass if submergence has taken place, which 

 is inevitable at the beginning if the rock has acquired, to any 

 considerable degree, the momentum of falling. 



The mass of rock, in the first place, tends to float upon the 

 lake of lava as a solid upon a liquid by which it is not wet, i. e. 

 by reason of the latter's surface tension. At Etna the writer 

 observed that huge bowlders which were incandescent and 

 which therefore amalgamated, if the expression may be used, 

 with the liquid lava, sank readily, while others, equally heavy 

 but cold and black, floated high and dry upon the liquid whose 

 surface tension was very great. 



The rock may also enter the lake at the place of a rising 

 lava current which will tend to buoy it up while, being cold 

 and of sufficient mass to retain its temperature for a time, the 

 lake material at once consolidates about it and forms a casing 

 and a fringe which are comparatively light and under which 

 the gases, constantly rising through the liquid, may accumu- 

 late and lend buo} T ancy to the whole. 



The rock itself has been long exposed to the atmosphere and 

 to moisture with consequent oxidation, etc., and when brought 



