STUDY XI. I97 



This would be the proper place to fpeak of the 

 roots of vegetables; but I am little acquainted 

 with what paries under ground. Befides, in all 

 Latitudes, on heights, as well as by the water's- 

 fide, we find the fame fubltances nearly, muds, 

 fands, pure mould, rock, which mult produce a 

 much greater refemblance in the roots of plants 

 than in the other parts of their vegetation. I have 

 no doubt, however, that Nature has eltablifhed, 

 on this fubjecl:, relations, the knowledge of which 

 would be highly ufeful, and that a cultivator, 

 fomewhat experienced, might be able, by infpect- 

 ing the root of a vegetable, to determine the fpe- 

 cies of foil belt adapted to it. Thofe which are 

 very hairy feem molt proper for fandy grounds. 

 The cocoa-tree, which grows to a very large fize 

 on the fhores of the Torrid Zone, thrives in pure 

 fand, which it interlaces with fuch a prodigious 

 quantity of hairy fibres, as to form a fol id 

 mafs around it. It is on this bafis that it effec- 

 tually refills the mofl violent tern pelts, in the 

 mid It of a moving foil. What is fingularly re- 

 markable in the cafe of this plant, it never fuc- 

 ceeds fo well as in the fand on the fea-fhore, and 

 generally languilhes in the interior of a country. 



The Maldivia Illands, which are, for the mod 

 part, nothing but fandy fhallows, are the molt re- 

 nowned regions of all Afia, for the abundance and 



o 3 the 



