474 



Mr. S. G. Starling on the Equilibrium oj 



soft iron bars, placed under or over the compass-bowl and 

 hinged together (fig. 11), and so arranged that a cord would 



Fig. 11. 

 V3> 



raise the ends of the bars N S as the tilt increases. As no 

 measurements are available, it is not possible to say that the 

 raising and lowering could be made to give a correcting 

 field which would vary suitably with the tilt; but the 

 correction would be of the right sign and would approximate 

 to a sine curve as the course changes. As the curves of 

 quadrantal deviation are somewhat similar to sine curves, it 

 might be possible to obtain an approximate correction in 

 this way. 



In the event of a quadrantal correction of this kind being 

 too small to be effective, it is also suggested that a pair of 

 soft iron blocks placed one on either side of the compass-bowl 

 would, by the attraction between the needle and them, 

 produce a deviation of the right character. It would thus 

 be necessary to move them nearer to the bowl as the machine 

 tilts, the distance away when the machine is level being so 

 great that they do not appreciably affect the position of the 

 compass-card. 



Measurements on corrector of type fig. 7.— Two vertical 

 magnets, one on either side of the compass-bowl, are carried 

 by an aluminium bar (fig. 12). The whole can rotate about 



