234 cook's voyage to feb. 



The observations for the variation of the compass 

 did not agree very well among themselves. It is 

 true, they were not all made exactly in the same spot. 

 The different situations, however, could make very 

 little difference. But the whole will be best seen by 

 casting an eye on the following table. 



East Mean 



Time. Latitude. Longitude. Compass. Variation. Variation. 



Jan. (Gregory's 10 10' 10') 



18th. A.M. 21 12' 200 41' J Knight's 9 20 5 [-9 51' 38" 



( Martin's 10 4 40 ) 



19th. P.M. 21 51 200 20 I mght ' S , !? J JJJ I 10 37 20 



(Gregorys 11 12 30 J 



(Gregory's 9 1 20 

 28th. A.M. 21 22 199 56 \ Knight's 9 1 25 [- 9 26 57 



(Martin's 10 18 5 ) 



(Gregory's 11 21 15 ) 

 28th. P.M. 2136 199 50 1 Knight's 10 40 VI! 12 50 



(Martin's 11 37 50 ) 



Means of \ g ^ ^ ] _ io 17 11 



the above J j 



On Jan. 18th.21 12 200 41 the north end of 



the needle dipped 42 1' 7". 



The tides at the Sandwich Islands are so incon- 

 siderable, that with the great surf which broke against 

 the shore, it was hardly possible to tell at any time 

 whether we had high or low water, or whether it 

 ebbed or flowed. On the south side of Atooi, we 

 generally found a current setting to the westward or 

 north-westward ; but when we were at anchor off 

 Oneeheow, the current set nearly N. W. and S. E., 

 six hours one way, and six the other, and so 

 strong as to make the ships tend, though the wind 

 blew fresh. This was certainly a regular tide, and 

 as far as I could judge, the flood came from the 

 N.W. 



I now return to the progress of our voyage. On 

 the 7th, being in the latitude of 29 N., and in the 

 longitude of 200 E. } the wind veered to S. E. This 

 enabled us to steer N. E and E. \ which course we 



