262 cook's second voyage july, 



Longitude by observation 28° 38' 5&" 



Ditto, by the watch - : 28 55 45 



Error of the watch on our > __ ^ ^g x 

 arrival at Portsmouth - ) 



True longitude by the watch 28 39 18* 



I found the variation of the compass, by several 

 azimuths, taken by different compasses on board the 

 ship, to agree very well with the like observations 

 made by Mr. Wales on shore ; and yet the variation 

 thus found, is greater by 5°, than we found it to be 

 at sea ; for the azimuths taken on board, the evening 

 before we came into the bay, gave no more than 

 16° 18' west variation, and the evening after we 

 came out, 17° S3' west, 



I shall now give some account of the variation, as 

 observed in our run from the Island of Fernando 

 De Noronho, to Fayal. The least variation we found 

 was 37' W., which was the day after we left Fernando 

 De Noronho, and in the latitude of 33' S., longitude 

 32° 16' W. The next day, being nearly in the same 

 longitude, and in the latitude of 1°2«5'N. it w T as 

 1° 23' west ; and we did not find it increase till we 

 got into the latitude of 5° N., longitude 31° west. 

 After this our compasses gave different variations, viz. 

 from 3° 57 to 5° 11' W. till we arrived in the lati- 

 tude of 26° 44' north, longitude 41° west, when we 

 found 6° west. It then increased gradually, so that 

 in the latitude of 35° N. longitude 40° W., it was 

 10° 24' W. ; in the latitude of 38° 12' N., longitude 

 32£° W., it was 14° 47'; and in sight of Fayal 1 6° 18' 

 W., as mentioned above. 



Having left the bay at four in the morning of the 

 19th, I steered for the west end of St. George's Is- 

 land. As soon as we had passed it, I steered E. •£ S. 

 for the Island of Tercera j and after having run thir- 



16 



