128 
LETTERS FROM HIGH LATITUDES. 
ing to the sky during an eruption, are nothing but the 
superincumbent mass of waters in the pipe driven up in 
confusion before the steam at the moment it obtains its 
liberation. 1 
The accompanying sketch may perhaps help you to 
understand my meaning. 
1 Professor Bunsen has lately announced a chemical theory, which, 
I believe, has been received with favour by the scientific world. 
He points to the fact that water, after being long subjected to heat, 
loses much of the air contained in it, has the cohesion of its mole¬ 
cules much increased, and requires a higher temperature to bring 
it to the boil; at which moment the production of vapour be¬ 
comes so great, and so instantaneous, as to cause explosion. The 
bursting of furnace boilers is often attributable to this cause. Now, 
