XXXIV INTRODUCTIOfT, 



well acquainted with the force neceSary to 

 break a main topmaft when ftruck; they 

 will not, therefore, think that I exaggerate 

 %vhen I eftimate the velocity of the wind, 

 during its moft violent gufts, at an hundred 

 and fifty feet per fecond* The extraor- 

 dinary variation of the barometer, on fliores 

 fituated between the two tropics, is the only 

 index hitherto known, by which a hurricane 

 can be forefeen fome hours before its com- 

 mencement. 



At the time of the hurricane^ which took 

 place in the month of February 1771, the 

 fudden falling of the mercury gave me, as 

 well as M. Poivre, great uneafmefs. It 

 was then four in the afternoon, and M. 

 Poivre invited the port captain to his houfe; 

 but that officer, who had been an eye-witnefs 

 of the hurricane in the year 1761, was not 

 (Iruck, as we were, with the variation of 

 the baromcten He faid, that there were 



5 more 



