512 
DE8. T. ANDEESOX AND J. S. FLETT OX THE EEUETIOXS OF THE 
crater, while the amount ot solid matter dischaiged was very much smaller. Several 
possible explanations of this offer themselves. It may be that the velocity of the 
blast is greatest at that point where the slope of the hill gives place to the plain 
beneath, and the city stood exactly in that situation. It is more probable, however, 
that the avalanches of Pelee ai-e of a more mobile and fluid character than those of 
the Soufiiere, and hence acquire a greater velocity, and this must be a consequence 
of the slightly different ] )ro})ortions of gaseous and of solid matter which they contain. 
To judge by the crystalline state of the dust, as revealed by the microscope, there 
can have been no very great difference in the temperatures of the two magmas at the 
time when they were shivered into dust. That of Pelee contains rather more 
amorphous, glassy matter, and may have been slightly the hotter of the two, but not 
to any great degree. We may suppose that the percentage of gases in the dust 
cloud of Pelee was greater than in that of the Souffiere, the temperature at least 
equally high, the interfusion and admixture of the comj^onents equally perfect. The 
one may be compared to a heavy fluid, extremely mobile, the other to one still 
heavier, but more viscous. It would seem that in these circumstances, as the slopes 
down wdiich they sped were about equally steep, the former would acquire a greater 
velocity after a certain specified period, or after covering a cei'tain distance from the 
crater. It must also be kept in mind that the avalanche emitted by Montague Pelee 
on May 8th was confined to a very limited space, and that the north end of St. Pierre 
stands, apparently, not far from the centre of its path. 
This brings into prominence one property in wdiich tliese discharges differ from 
ordinary avalanches. They have, in virtue of the expanding gases with which they 
are permeated, the powder of moving down slojies much gentler than those on which 
an avalanche coidd start. The sides of Pelee and the Souffiere have an average 
inclination of 12 to 15 degrees, if we neglect the deep, steep-sided ravines with which 
they are seamed, and no avalanche, even of fine snow, would move on such gentle 
gradients. Had the hot sand been piled up there it would have rested peaceably; 
without tlie surging gases intermingled wdth the solid particles there would have 
l)een no motion. It is iK)t correct to regard these cataracts of sand as mere 
avalanches ; the idea of a blast is also essential, if Ave are to form a proper conception 
of the mechanism in operation. 
Several other explanations have been advanced to account for the manner in AAdiich 
the deadly cloud of the Pelean eruption SAA'eeps ovei- the ground, but does not 
ascend in the air like the ordinary steam clouds Avith Avhich all geologists are familiar 
in volcanic outljursts. One of these Ave may call the hypothesis of oblique discharges, 
the other the explosion hypothesis. It has also been hinted that electricity is 
responsible in some way or other for the deAmstation, but Ave are not aA\-are that 
anyone has as 3mt formulated any Avorkable hj^pothesis on this liasis. 
The great V-shaped fissure AAdiich looks doAvn on Sf. Pierre from near the summit of 
Pelee is undoubtedly responsible, l)y its disposition and conformation, for the course 
