386 
DRS. T. ANDERSON AND J. S. ELETT ON THE ERUPTIONS OF THE 
they saw the water of the lake, hut this is more than doubtful, others that they 
saw “ where the water run over.” The top of the hill was in a continual tremor. 
They turned and fled back along the path, and, meeting others ascending, told them 
what they had seen and all returned together. As they passed the estates by the 
roadside they spread the news that tlie Soufriere was in eruption, and the ■ lake had 
boiled over. They were received with incredulity, and when they came to George¬ 
town they were scofted at as fools and cowards. 
Discharge of the Crater Lake .—It is certain that the lake was boiling, though at its 
usual level, and in no way much altered on Tuesday at mid-day. That afternoon it 
was reported at Chateaubelair that the water was at the top of the lip and running 
over. This statement we were not able to confirm by evidence collected from eye¬ 
witnesses, but no serious discharge can have taken place that day or the following 
night. Mr. Robertson was at Wallibu, and the valleys on the windward side had 
]louses standing on tlie banks of the streams and full of people. The lake was a 
quarter of a square fhile in area and at least 500 feet deep. Its discharge would 
have produced tremendous floods of boiling water, much greater than the heaviest 
of tropical rains could ever have occasioned, and such floods cpuld not possibly have 
taken place without being noticed. Tlie natural channel of escape for the lake was 
over the lower or southern lip and down the valleys of these rivei'S, and we may be 
sure the bulk of tlie water had not been discharged before Tuesday morning. 
It is in every way proliable that the first discharges of the crater lake took place 
during the night of Tuesday. Mr. Morgan, of Chateaubelair, told us that early on 
Wednesday morning large trees were seen floating in the mouth of the bay on which 
that village stands, “as if there liad been heavy floods in the Larikai River.” They 
were uprooted and had lost their branches, but were not burnt, and there had been 
no heavy showers during the night. Probably gushes of water liad flowed down that 
valley or the ravine of the Rozeau Dry River, and swept before them some of the 
trees growing on the slopes. We must remember, also, that the fish sellers who 
were near the lip of the crater on Wednesday, at about 9 o’clock, all agreed in 
stating that they saw where “ the water run over,” and that the lip of the crater was 
covered with soft mud through which they hesitated to walk to where the path 
descends to leeward. About 11 o’clock, Mrs. Kelly, of Wallibu, who was in a boat 
ofl‘ the mouth of the Wallibu River, saw “ an upheaval of vapour coming from the 
Wallibu Dry River, just at the beach.” This is the first mention of hot water in any 
of the stream coiirses. 
Of the first explosion after day-break, that at 6 A.M., Mr. McDonald states that 
Ijelow “ the usual tall, column of white vapour ” there was “ a much shorter column of 
dense black stufl’ whicli seemed heavier, as it quickly subsided back into the crater. 
This was the first appearance noted of probable solid matter being ejected.” It may 
have been a column of mud and muddy water mixed with stones. The fish sellers, 
a couple of hours later, found stones lying on the road leading up Ifom the windward 
