DIFFERENCES IN THE MAGNETIC NEEDLE. 103 



It is apparent that fome of the obferved variations in the The errors weie 

 above table are 4» lefs and others 4° greater than the truth ; ^"^""^ 4° each 



" _ _ way, and the 



and it maybe remarked, that when this error is weft ward, north end of the 

 the fliip's head was eaft, or nearly fo, and when it was eaft- "'^^^t's deviated 



3S it rcDcllcu bv 



ward the head was in the oppofite diredion. When the the fhig's head, 



obrervatioiis agree neareft with what was taken on Qiore, or ^'^• 



with what may be deemed the true variations the fliip's head 



was nearly north or fouth ; and a minute infpe6lion of th^ 



(able will favour the opinion, that the excefs or diminution of 



the variation was generally in proportion as the (hip's head 



inclined on either fide from the magnetic meridian. 



After I had well afcertained the cerlainty of a difference in 

 the compaffes, arifing from an alteration in the point (leered, 

 I judged it neceflary, when I wanted a fet of bearings from a 

 point where we tacked the fliip, to take one fetjuft before and 

 another immediately after that operation ; fome fpecimens of 

 thefe here follow. 



1 802. Head ESE, Head SW b. W. Other obfcrva- 



April 13th, rLeGeographe *'""*• 



I Rocks - N55°to7I«E 

 11*'32'AM I H point - N 4 W after tacking N 9^ W 



— In point - S 32 E - S 40 E. 



HeadSEKY.. IleadW. 



April 14th, I* n point rocky, 



J inner part N 39® E. after tacking N 30' E 

 9^ 29' AM 1 pro- 



L jeding part N 67 E - N 59 E 



— Furtheft vifible 



extreme from 



deck - S 51 E - S 55 E. 



Head ENE. Head SW b. S. 



April 15th, rn, the wefterri 



J part - Nl5<>W. after tacking N 21 oW 

 11'' 50' AM I A peaked hum- 



L mock - Nl9 E - Nl5 E 



— Furtheft extreme 



from deck S 53 E - S 61 E 



— Centre of a naked 



fandy patch - E - E 5 N. 



Variation per amplitude April") 



15, AM, taken with the fur- > 4» S' E, fliip's head being S# 

 veying conopafs 3 



April 



