VOLCANIC ERUPTIONS ON MARTINIQUE AND ST. VINCENT. ^-545 



but are .similar to those occasioually ticcompanyinj^' earth(iiiake.s. So 

 far as can l)e judged, however, the unusual waves that have recently 

 broken on the shores of Martinique^ and St. Vincent have not ])een due 

 to movements in the earth's crust, sui'h as commonly produce earth- 

 quakes, although some of them may have been of that nature. The 

 waves refei'red to have been caused, in most instances, by the distur!)- 

 ances produced in the water of the sea l)v the l)lasts of dust-laden 

 steam that have swept down from the craters of Mont Pelee and La 

 Soufriere. Similar waves ha\-e also ])een generated h.y the entrance 

 into the sea of stupendous nuid Hows, or, perhaps more properly, 

 avalanches of rock debris and water, like the one which destroyed the 

 Guerin sugar factory on May 5. Again, landslides lia\'e occurred in 

 the loose dejiosits on the Caribbean shores of both Martini([ue and St. 

 Vincent, and similar slides, as indicated by the ))reaking of telegraph 

 cables, have probably taken place on the steep submerged slopes of the 

 mountains whose sununits form the islands mentioned. In these several 

 ways waves in the sea appear to have been generated, ])ut in all instances 

 they have been low and but little damage from them has resulted. 

 The earthquake shocks that accompanied (he recent ei-uj)ti()ns have 

 been comparatividy light, and, so far as can be judged, not of such a 

 nature as to cause large; waves in the adjacent sea. The earthcjuake 

 shocks, however, may and probably did ])ring about the descent of 

 some of the landslides on the margin of the sea and on the steep sub- 

 merged slopes, and in this way are indirectly accountable for some of 

 the sea waves. 



La/idslides. — The landslides just referred to occurred principally on 

 the west side of St. Vincent, to the north of Chateaulxdair, where 

 strips of nearly flat alluvial land, adjacent to the sea, have disapp(>ared, 

 leaving fresh blulfs of loose deV)ris some Ho or 40 feet high. It has 

 been suggested that this disappearance of land, ajul in one instance 

 of the site of a village, was due to moviMuent along a fault — that is, 

 the subsidence of the rocks on one side of a deep fracture in the 

 earth's crust — Init the evidence does not seem to sustain this hypoth- 

 esis. The lands that have disappeared, as shown by the escarpments 

 remaining, were composed of unconsolidated de])ris, deposited for the 

 most part directly by streams as deltas. This loose matiu-ial, resting 

 on the steeply inclined rocks beneath, Avas in a })osition to be easily 

 dislodged bj- earthquake shocks, by the rush of nnid avalanches down 

 the valley, at the mouths of which the deposits had btM>n made, and by 

 the return waves when the sea was disturl)cd by the volcanic Idasts or 

 by mud avalanches. The changes made by subsidence of the land are 

 not great, and, as several observers have stated, may reasonably be 

 accounted for in the manner just referred to. 



