Magnetic Needle as observed on Skip-loard. 283 



Transactions, suggested that they were probably owing to the 

 concentration of the magnetic influence of the iron made use of 

 in the construction of the ship. The truth of this suggestion, and 

 the accuracy of his observations, have since met with full con- 

 firmation, and his practical rules founded thereon have received 

 additional support, from the " Essay" of Mr. Bain " on the Va- 

 riation of the Compass," published last year. 



As I have been materially anticipated by Mr. Bain in a series 

 of observations on the variation of the compass *, which 1 con- 

 ducted on the coast of Spitzbergen, in the years 1815 and 1817, 

 it will be unnecessary here to enter into the detail of these ob- 

 servations, or enlarge upon the probable cause of the anomalies 

 observed ; it may be sufficient to give a table of the most accu- 

 rate of my observations, and annex to it the few general infe- 

 rences which were drawn from it, during the voyage in which 

 most of the observations were made, together with such remarks 

 on each inference as seemed to me calculated for its elucidation. 

 I shall however just premise, that I am not unconscious of the 

 great liability to error in observations of this kind, and of the va- 

 riety of causes (arising out of the unequal distribution of iron in 

 different ships, whereby numerous local attractions are formed) 

 which contribute to the multiplication of those errors: it is, there- 

 fore, with the greatest deference that I submit these deductions, 

 particularly as I conceive it will require observations to be made 

 tinder a vast variety of circumstances, and in many different ves- 

 sels, before correct and satisfactory conclusions can be drawn. 

 It is 07ily then as a step towards facilitating such general con- 

 clusions, the importance of which to our maritime concerns is so 

 obvious, that I presume to offer these observations and remarks. 



• The azimuths, contained in the following table were talien, either by the 

 needle of a theodolite, or by a compass fitted up at sea, for the purpose, with 

 a card made extremely light, and a bar fastened edgewise to it, by two brass 

 screws, a a, as in the annexed sketch. The compass being small, the card 

 light, and the needle very powerful, owing to the tlnckness of its ends f, it 

 performed consideral)lyl)etter than an expensive azimuth compass of larger 

 dimensions, which indeed was so sluggish and erroneous in its indications, 

 that I could make no good use of it. 



t See Plate IV. fig. 3. 



Tahle 



