Xlll 



attention of the Committee. Smith's attention was called to the subject 

 by his friend Sabine ; and he gave a rule for placing the needles, which 

 was adopted by the Committee, and has ever since been followed in the 

 construction of the Admiralty compass. The rule is, that when there are 

 two needles used they should be placed with their ends on the compass 

 card at 60 on each side of the ends of a diameter ; and that when (as 

 in the Admiralty Standard Compass) there are four needles, they should 

 be placed with their ends at 15 and 45 from the ends of tke diameter. 

 The object of this rule was to give equal moments of inertia round all 

 horizontal axes, and so to remedy the <k wabbling" motion of the compass 

 card when balanced on its pivot, which has been found inconvenient. 

 Captain Evans, in a letter recently received from him by the writer of 

 this notice, says that the " wabbling " motion has been satisfactorily 

 corrected by this arrangement of needles ; " it is transformed into . a 

 * swimming ' motion;' 



About twenty years later it w r as discovered that the same arrangement 

 gives, by a happy coincidence, a very important magnetic merit to the 

 Admiralty compass, which had not been contemplated by Smith when he 

 first gave his rule. To explain this, it must be premised that practical 

 compass-adjusters had experienced difficulties in correcting the compass 

 deviation of certain ships by Airy's method (which consists in using soft 

 iron to correct the quadrantal deviation, and permanent magnets to 

 correct the semicircular), and had reported that in such cases they had 

 found it advantageous to substitute compasses with two needles for a 

 single-needle compass. The attention of Captain Evans was drawn to 

 this subject by the observations made in the * Great Eastern ' on her 

 experimental voyage from the Thames to Portland, and afterwards when 

 she was lying at Holyhead and Southampton, from which he found that 

 although the deviations had been carefully corrected by Mr. Gray, of 

 Liverpool, with magnets and soft iron, and were in fact nearly correct 

 oil the cardinal and quadrantal points, there were errors of between 

 5 and 6 on some of the intermediate points. These observations indi- 

 cated the existence of a considerable error, which was neither semicircular 

 nor quadrantal, and thus apparently some source of error which had not 

 been taken into account by Airy in his plan for correction. To explain 

 the cause of these and similar results in other ships, previously considered 

 to be anomalous, Captain Evans instituted a series of experiments with 

 compasses, and magnets and soft iron placed in different positions with 

 respect to them. He soon found that the greatness of the supposed 

 anomaly in the ' Great Eastern ' depended on the unusually great length 

 of the needles of her standard compass (two needles* of 11| inches in 



* Compass needles becoming larger with the ships, by a process of "Artificial Selec- 

 tion " unguided by intelligence, have sometimes attained to the monstrous length of 

 15 inches, or even more, in some of the great modern passenger-steamers fitted out by 

 owners regardless of expense, and only desiring efficiency, trusting to instrument-makers 



