Ch. IV. SOUTH AMERICA. 341 



rled us northward. On the 17th they continued the 

 fame courfe, but on the iSth we found 27 min. dif- 

 ference i having fuddenly entered into writers of a 

 different courfe, that is, they again carried us fouth- 

 ward. Had this variety of difícrcnces been reported 

 only by a fingle perfon, doubtlefs from the ftrange- 

 nefs of the plisnomenon, it would have beelt looked 

 upon as a miilake in the obfervation ; but the 

 reality of thefe dangers does not now admit of 

 any rational doubt. The obfervations were taken by 

 ieyen perfons, each with a different inftrument, one 

 of which was a quadrant invented by the celebrated 

 Mr. Hadley, and all agreed in the diaerences. From 

 the i 8th day of the month to the 20th the diffe- 

 rence betwixt the latitudes by obfervation, and that 

 by account, proved to be 40 min. which was near 

 double of that which had appeared in the courfe of 

 the i8ch ; and on tlie aoth we were in the latitude of 

 43 deg. )t min. and 38 deg. 57 min. E. of the meri- 

 dian of Conception. 



Notice has already been taken of the indifferent 

 condition of the Delivrance at our departure from 

 the ifland of Fernando de No!-ona. And thus it 

 continued dll the i 6th, when whether it proceeded 

 from the working of the fhip or any other caufe, 

 the water increafed fo fail that the fhip was reduced 

 to that diftrefs, which had been the chief occafion 

 of our putting into the above ifland, and of our 

 long Hay there. On the aoth it increafed upon us 

 •in luch a manner that the pumps were kept conti- 

 nually going the whole night; but on the 21ft it 

 fuddenly abated to near a fourth part of what it 

 had been the day before ; a moil fealbnable relief to 

 the fiiip's company fpent with a fatigue, whic!i their 

 fears rendered ftill more painful. We obierved that 

 the water continued to dccreafe from the dme of 

 our getting into the fargafo or weeds, fo that on the 

 ayth the fhip fcarce made an eighth part of the water 

 '^3 as 



