324 



« 



particularly the village, where they have endeavoured to 

 procure a supply by sinking wells in the coral-bank; but 

 the produce, though fit for use, is rather brackish. 



" We left St. Leu early in the morning of the 14th, and 

 pursued our journey over a road encumbered with stones, 

 but nearly destitute of those ravines which had rendered our 

 progress during the preceding stage so slow and difficult. 

 The land is cleared and cultivated to a great height along 

 the acclivity of the mountain ; and the only memorials re- 

 maining of its former state, are a few Lataniers and Benzoin 

 trees, battered by the winds, and verging gradually to decay. 

 About twelve miles from St. Leu, we passed a remarkable 

 sand-bank, which extends upwards of a mile in length, and 

 seems to be creeping slowly up the side of the mountain. 

 The sand is of a bluish-grey colour, and affects the form of 

 ridges or wreaths, in the manner of drifted snow. It might 

 almost be supposed that some extraordinary cause had 

 accumulated all the sand in the island on this particular 

 spot, as there is hardly a particle to be seen in any other place. 



" The church of St. Louis, for there is no village in the 

 district of that name, stands about four miles from the sand- 

 bank. We purposed halting here for a night, and called on 

 a planter to whom we had letters of introduction ; but he 

 happened to be from home, and we agreed to prolong our 

 ride to St. Pierre. Close by St. Louis we crossed the River 

 St. Etienne, at this time an inconsiderable stream, but with 

 a cliannel several hundred yards wide, and obstructed by 

 huge masses of stone jammed against each other. Within 

 two miles of St. Pierre, we passed over a high rampart of 

 lava, the uncommon aspect of which excited our attention. 

 It is a confused mass of gravel, pebbles, and angular 

 fragments of trap, firmly agglutinated by a cement of the 

 appearance of mud, and differing very little, on a superficial 

 view, from conglomerate. As this rampart is composed of 

 several distinct beds lying over each other, with a surface 

 nearly destitute of vegetation, it may be reasonably inferred 

 that it is the result of successive and comparatively recent 

 eruptions. 



