G. P. Scrope on Craters, and the Liquidity of Lavas. 355 
upwards, and after bursting like a great bubble, fall back again 
out of sight. In 1819 I was myself able to witness the same 
interesting phenomenon probably from the same position, a high 
point of the external crater-rim’ which overlooks the vent. At 
each belch, a shower of tattered fragments of lava, torn from 
the surface of the bubble as it broke, rose into the air with a 
cloud of vapor and a fierce roar; while steam seemed to be at 
intervals blowing off from another neighboring vent. Hoffman, 
who visited the same voleano a few years later, describes in mi- 
nute detail precisely the same phenomena. 
The vast size of some craters, already noticed, may afford a 
notion of the enormous volumes of gaselform matter that must 
ave been discharged through them at the time of their forma- 
tion by continuous explosions lasting for weeks and even months; 
Since each individual bubble of vapor must have been of a mag- 
nitude to fill the entire horizontal section of the crater; and 
even for some time to aid in enlarging the area of this aperture 
by violent pressure against its rocky sides. The prodigious 
force with which they ascend, and therefore the great depth at 
which they are generated, may be judged from the vast vertical 
ht, measured in miles, to which they have been seen to 
shoot up a continuous columnar fountain of ejections, cons ~~ 
ts 
i 
not merely of scoria and ashes, but often of rocky fragmen 
great size. 
Sssure, which in the preceding year had been visibly broken 
through the side of the cone towards the northeast. Sometimes 
