APPENDIX. 159 



ward ; at the same time it is difficult to account for n.e. by 

 compass being n.e. true, as appears to have been the case on 

 the 19th, because the longitude being still very nearly the 

 same, there is no reason to apprehend an increase of actual 

 westerly variation. 



(71) The still farther advance to the southward, and the 

 consequent diminution of dip and increase in the terrestrial 

 intensity, is sufficiently seen in these observations, although, 

 with the ship's head to the southward, the needle is still 

 weak and indeterminate in its directive quality. 



(0) (p) The vessel being now returning towards the station 

 where the variation had been found to be 37|o ^v.. but to 

 the eastward of it there can be little doubt that the varia- 

 tion was from the 23d to the 25th, about 3i or 4 points 

 west ; taking the least, the local attraction on the 23d 

 would be 3|- points, with the head at east by compass, and 

 21- points on the 24th, with the head at n.e. ; about 33° 

 atN.E.b.E., and 36° at w.b.N. on the 2oth. That is, by sub- 

 tracting 7° 27' from the variation in the one case, and in 

 the other subtracting the variation from 75° 49' w., which 

 results are exactly such as we should have naturally antici- 

 cipated. 



(^) The irregularities in these deviations show that, with 

 the ship's head to the southward, the compass had not yet 

 recovered its proper action. 



(/•) These results indicate very clearly a variation of 

 4 points westerly, and a local attraction of 2 points at east 

 and west. 



(s) The mean of the two variations at east and west seems 

 to indicate here a westerly variation, amounting to 5^ points 

 and a local attraction of about one point and a half; that 

 is, about double the attraction in England : both of which 

 indications are supported by a high degree of probability. 



