472 
DES. T. AXDEESOX AXD J. S. FLEET OX THE EEUPTIOXS OF THE 
levels at v'hich the stream flovecl as it gradually eroded the mass vhich had 
obstructed its older and still wider bed. The “lava bed” is composed of black 
mud, sand, and stones, in which there is also a considerable amount of charred 
timber. It is essentially a recent formation, which has filled up and obliterated the 
course of the stream at no very distant date, as it differs considerably in appearance 
from tlie rocks of the Carib Country, in which the primitive valley of the llabaka 
Paver had been carved, and has altogether a much newer aspect than the weathered 
tufis exposed on the sea cliffs and in the higher slopes and ravines. Many beheve 
that it is a product of the eruption of 1812, but we could find no very conclusive 
evidence of this, and there is no inherent reason whv it mav not have been due to 
some earlier eruption. 
It faces the river in low. vertical banks from 6 feet to 20 feet high ; its upper 
sui'face is perfectly terraced, and shows two or three benches several feet apart. 
Only two explanations suggest themselves to account for its presence there. It 
mav have been a hot sand avalanche, and, if so, it cannot have accumulated during 
the eruption of 1812, for when so great a mass of debris was projected mto the 
lower part of the Pabaka Valley the accompanying hot blast must have been 
tremendous, and would have annihilated every living thing for miles around. But 
it does not look like a sand avalanche; in the stream sections it too often 
shows a V ell-marked bedding, and it contains many rounded blocks of lava, which 
certainlv appeared to be water-woin boulders, such as are common in the valleys. 
Evervthing pointed ratlier to its having been a great mud lava, which had swept 
down the upper parts of the river’s course with a high velocity, and had caught up 
and incorjDorated the gravel and boukPrs over which it passed. Then, when it reached 
the fiat country at the lower end of the valley, it had been unable to flow further, and 
had come to rest, a great glacier of black mud and stones, which filled up the broad but 
shallow channel into which it had flowed. Such floods of mud followed the eruption 
of 1812, and may possibly occur in the near future as a consequence of the eruption 
of this vear, and the discharo’e of snch mud lakes as are formino; in some of the 
hio’her sti'eams would furnish most of the conditions necessarv for the obstruction of 
the livers in their lower parts by masses of stony mud exactly resembling the 
“ lava bed.” 
The ''May Dust'’ in Barhados, 1812. 
Earlv in the morning'' of Mav 1st, 1812, sounds as of distant cannonadino- were 
heard in Barbados, and it was generally believed that a naval engagement was 
taking place somewhere off the coast. The garrison prepared to I'epel any attack. 
At Bozeau, m Dominica, similar noises were heard a little after midnight, and the 
regular forces were placed under arms and the militia called out. In Barbados dust 
began to fall about half past 1 o’clock, and there was intense darkness, but in 
Dominica there was no darkness and no fall of dust. For several days afterwards 
