331 



excavated in the sand, within three or four miles of the sea. 

 Such springs may be looked for with a certainty of finding 

 them round the whole coast. The known laws of hydrostatics 

 would indeed teach us to expect water in all such situations, 

 not only in this island but in every part of the globe. 



" We left La Renaudie's about sunrise, and shortly 

 thereafter crossed the River de I'Este, which bounds the 

 district of Ste. Rose on the north. This is the largest and 

 most impetuous torrent in the whole island. Every other 

 river lias its periods of quiescence ; this is perpetually in an 

 uproar. Even at the time when we crossed it, though the 

 stream was hardly five yards over^ the passage was not effected 

 without some risk, its depth and rapidity rendering it always 

 unsafe. Its winter channel is a quarter of a mile in breadth, 

 and paved with enormous masses of stone rolled down from 

 the mountain, and wedged against each other. 



" After crossing the river, the road leads along a belt of 

 alluvial land close by the shore. The acclivity of the moun- 

 tain on the left hand is cleared to a great height, and covered 

 with flourishing plantations. This district enjoys, in common 

 with Ste. Rose, the benefit of copious showers from the east- 

 ward, which enhance its fertility. After an easy ride, we 

 arrived in the evening at the habitation of M. Hubert, where 

 we took up our lodgings for the night. Next morning we 

 breakfasted at the house of M. de Jean, in the district of Ste. 

 Susanne, and afterwards walked through his garden, where 

 I for the first time saw the Garcinia Mangostana and G. Gam- 

 bogia. The Litchi trees [Dimocarpus Litchi) in this garden 

 were so loaded as to render it necessary to have their branches 

 propped, to prevent their breaking down under the weight 

 of the fruit. The nature of this fruit does not seem to be 

 well understood by Botanists. In all our systematical works 

 it is termed a berry; but it is in reality a two-valved cap- 

 sule, (the second germen proving always abortive,) containing 

 a single large seed, or rather nut, invested with a fleshy 

 arillus. This last, which is the only eatable part of the fruit, 

 is entirely unconnected either with the nut or the capsule, 

 except at the very base. It is open and jagged at the top, 



