164 COMPENDIUM OF GEOGRAPHY AND TRAVEL 
minutes, increased in violence, until at night there was 
an almost continuous roar. The sounds thus caused 
were indescribable, as may indeed be easily conceived, 
seeing that we have only the discharge of heavy cannon 
or thunder to serve as means of comparison. By one 
observer they were stated to resemble discharges of 
artillery at every second of time, combined with a 
crackling noise which was probably due to the collision 
of the fragments of the ejecta. But at Buitenzorg, 100 
miles distant, a similar comparison was used, and the 
noise compared to the firing of a park of artillery close at 
hand,—so violent, indeed, were the explosions that the 
windows w T ere blown in and sleep was rendered almost 
impossible. 
Such were the phenomena witnessed by the onlookers 
who escaped with their lives from the horrors of the night 
of August 26 th. Precisely what occurred at the focus 
of eruption will always remain a matter of doubt, but it is 
probable that from long-continued eruptive action the lip 
of the crater became gradually removed, and the sea 
gained admission to the white-hot mass of lava in the 
interior. The extraordinary violence of the explosions 
may have been due to the immense amount of steam thus 
generated, or the inrush of a vast body of water may have 
had a merely mechanical action, blocking the vent like 
the clods of earth thrown in to cause the eruption of a 
geyser. It is not unlikely that the succession of smaller 
seismic waves which left the island at various times during 
the night of the 26 th, owed their origin to these causes, 
but the great wave, which will be presently referred to, 
and which was productive of the most wide-spread results, 
is considered by most authorities to have been aided, if 
not entirely caused, by some upheaval of the sea-bottom. 
It is worthy of remark, however, that at no time and 
