APP. N° IX. J MAGNKTICAL OBSKRVATIOXS. 553 



section of the line joining the two situations in opposite sides 

 of the ship, with the midship hne traced fore and aft, will 

 probably point out a situation directly over the top of the focus 

 of attraction, where no anomaly on any course whatever will 

 appear. 



a. The Esk, in which I made my magnetical observations, 

 had, as above stated, an armourer"'s forge near the lar- 

 board bow, which, with the varying position of large 

 quantities of iron-work, composing our whale-fishing 

 apparatus, contributed to vary this point, where no 

 anomaly is supposed to exist, and prevented me from 

 ascertaining satisfactorily, at any time, its precise si- 

 tuation. I made, indeed, but few observations with 

 this view, and these, I find, neither establish nor re- 

 fute the inference. 



12. The anomaly of variation is probably the greatest in 

 men of war, and in ships which contain large quantities of 

 iron ; but it exists in a very considerable degree also in mer- 

 chantmen, where iron forms no part of the cargo, especially 

 in high latitudes, where the dip of the needle is great. 



(a.) A model of a vessel built of timber and plank, and fas- 

 tened with iron, which I have procured for showing 

 the nature of the magnetic anomaly, as connected with 

 tlie course steered and the position of the compass, is 

 capable of illustrating almost the whole of the preceding 

 inferences. In this model, as well as in actual prac- 

 tice, the vertical iron only is found to be capable of 

 affecting the indications of the compass ; the largest 

 pieces of iron placed horizontally, unless within an inch 

 of the needle, having little or no influence, while 

 the same, in a vertical position, produce a sensible 

 effect at the distance of almost a foot. 



(6.) While I was copying this paper for the press, a little 

 model of a vessel was shown me by Dr Traill, made 

 by a Mr Bywater of Liverpool, for exhibiting the 

 superior attraction of upright bars of iron over hori- 



