SOUFPJERE, AND ON A VISIT TO AfONTAONE PELEE, IN 1902. 427 
roads from tlie slopes behind. From these last the ash was mostly washed away and 
deep furrows had been cut in the naked soil, 
A few houses were set on fire, perhaps by lightning or possibly by hot falling stones. 
Others had collapsed under the weight of the ash which gathered on the roofs, while 
several had been knocked down by streams of mud on the day of the heavy rains. 
The difference between the thickness of the ash deposit and the amount of damage 
done in Georo-etown and at Chateaubelair is so strikinof as to deserve discussion. 
Though Georgetown suffered most, it is quite a mile more distant from the crater. 
There is no reason to believe that the black cloud of dust passed over either of these 
places; the effects observed are entirely due to the rain of ashes during the afternoon 
and night of the 7th May. Whatever may be the reason, it is clear that the crater 
projected a larger amount of material to eastward than to westward. This is in 
accordance with observations made by Mr. McDonald from Chateaubelair during the 
forenoon and up till 2 o’clock on Wednesday. He saw many showers of stones 
“ principally to windward.” Part of the dust, however, that fell in Georgetown 
may be due to a slow lateral spreading of the margins of the black cloud after its 
first velocity and heat had diminished. Chateaubelair is protected by several ridges 
and spurs which lie between it and the crater, but to the north of Georgetown sj^reads 
the level Carib Country. The strength of the trade-wind appears to have been unable 
to direct most of the material to leeward : it must have been hurled out of the throat 
of the volcano in an eastward direction against the wind. 
All the country north of Black Point (south of Georgetown) on the windward, and 
from the north side of Chateaubelair on the leeward coast, and of a line passing between 
these two points and over the summit of Morne Garu, showed on its surface a covering 
of ashes more or less deep. Within this area there was another less extensive region 
in which the damage to vegetation had been severe, and everything presented a burnt 
and scorched appearance. This is the true “ devastated country,” and its southern 
border starts about half a mile north of Georgetown, and running along the ridges to 
the summit of Morne Garu, descends by the spurs on the leeward side to the south side 
of the mouth of the Richmond River, a mile north of Chateaubelair. It does not 
include quite the whole of St. Vuncent north of this line : as at Owia and Sandy Bay, 
in the north-eastern corner of the island, there was comparatively little damage done 
to the vegetation (though the ash gathered to a depth of several inches). The boundary 
must be represented by a line drawn so as to exclude this quarter. (See map, 
Plate 39.) In the country included within this line not only was vegetation greatly 
injured or destroyed, but many of the inhabitants and most of the cattle in the fields 
were killed. Outside this line no one died directly from injuries received from the 
volcano. Consequently it marks the limit of the “ danger zone.” 
3 I 2 
