36 



Scientific Proceedings, Royal Dublin Society. 



It soon became evident that the method was susceptible of great 

 accuracy, and is yet capable of further improvement; but we content 

 ourselves with indicating the lines along which we worked, as the results 

 obtained were sufficient for our purpose, leaving other refinements to those 

 who may be induced to pursue the matter further with a view to obtain 

 quantitative results. 



N G 



©. 



©S 



: ■ : } :' :' i 



->// 



Fig. 1. 

 Errors of the Method. 



We have already referred to surface-tension and its practical elimination 

 However, floating dust, bubbles, &c, deviate the path of the float, and must 

 be avoided. Their effect is easily seen by the recorder, and when observed 

 he effaces the curve commenced, and repeats the record after the controller 

 has skimmed the peccant particle away. 



Fig. 1 shows the trajectories of the paraffin sphere, when started 

 from various points in the field of two unlike magnetic poles, which were 

 reversed with respect to the horizontal field of the earth's magnetism. 

 The figure bears a striking similarity to the well-known distribution of lines 

 of force for like poles. 



The cross + marks the point equidistant from either pole-piece, and 

 will be referred to as the Centre of the Field or C. F. The © marks the spot 

 where the sphere remains at rest. For a small displacement along the 

 diameter this is a position of stable equilibrium, but for a displacement 



