OF THE CARIBBEE ISLANDS. 75 



dies, such as the leeward shore of Guadaloupe, immediately 

 under the great SoufFriere, where hot boiling springs are found 

 within the high-water mark, or in the channel between the 

 volcanic promontories of St Kitt and St Eustatia, where the 

 fish are so remarkably poisonous, that fishermen rarely exer- 

 cise their trade in that part of the sea. It abounds, I have 

 heard, in a particular manner, with the yellow-billed sprats, 

 and no inhabitant will venture to eat a baracoota or other 

 large fish taken thereabout. 



Most of the accounts agreed in what I had myself obser- 

 ved, of the fish that proved poisonous being always in the ve- 

 ry best condition, and generally the largest of their species, 

 without marks of disease of any kind in any part of them, and 

 I never heard of a single exception to this, the sufferers ha- 

 ving always been induced to eat heartily, from the peculiar 

 excellence of the fish; and all who were capable of examining 

 the subject, were convinced of the inutility and frivolity of the 

 tests usually employed, such as boiling a piece of silver, &c. 

 with the fish, adducing instances where the severest poison 

 had been received when the test declared security, and vice 

 versa ; the only test hitherto discovered, in the least to be de- 

 pended on, being that of giving a portion of the fish, the offal 

 in. preference, to a cat, a duck, or a pig, and witnessing its ef- 

 fects upon them, before proceeding to make use of it. 



The commonly received opinion, too, of copper banks at the 

 bottom of the sea, that is to say, uncovered and insoluble cop- 

 per ores in contact with the waters, the very existence of 

 which is a most improbable gratuitous supposition, communi- 

 cating a poisonous quality, such as that found in the yellow- 

 billed sprat, to the wandering fishes that roam in the waters, 

 I need scarcely say to the Society, was refuted and shewn to 

 be impossible ; as well as another opinion, almost ?is general, 



K 2 that 



