300 Analyses of Books. [Oct. 



130. Accordins; to the hypothesis above adverted to, it is almost impossible to 

 sec beforehand wlietlier any given situatinii of the i)all will produce an easterly or 

 westerly deviation, and more particularly lo form any idea of the quantity, as to 

 greater or less. In all the early part of mv experiments, I had this dirticulty to 

 encounter ; and it was only after some reflection ihat I was enabled to arrange the 

 results in a manner which presented a certain degree of order : this order was ren- 

 dered, however, tolerably perspicuous after 1 had determined the existence of the 

 plane of no attr.iction. 



131. I had already arrived at this conclusion when I mentioned the results of mv 

 experiments to my colleague, S. 11. Christie, lisq. who became much interested in 

 the inquiry, and afterwards spent willi me some time in the pursuit; and to him I 

 owe the following simple idea of the mode of action, which, wliether it be actually 

 correct or not, will serve (as Dr. Young has remarked, with reference to the theory 

 of Coulomb), as a convenient vehicle of illustration, and for uniting under one gene- 

 ral head a number of facts which would be otlii'iw i?e detached and iiisi\lated. 



On examining the nature of the several resultsobtained in ijjy first series of experi- 

 ments, adverted to in Sect. I. (he above-named i;enileman observed, that holh the qua- 

 lity and quantity of the several deviations were (as nearly as could le tlien asc er- 

 tained) such as would liavc had place, supposing the needle to have been inclined 

 in its natural direction, and then referred to llie horizontal plane: in fact, that 

 ■whatever might he the posilion of the ball, the result still seemed to have a certain 

 relation with the quality and quantity of the dcviaiion which would take place on 

 a line passing through the centre of the needle, and having an inclination corres- 

 ponding to that of the dip. This line, for the pur|)ose of fixing the ideas, we may 

 imagine to be a fluid passing in this direction, and that the deviation is caused by 

 the action of the iron upon either of its branches to which the iron is most expos- 

 ed. Let us examine how far, by admitting this view of the subject, we may be 

 able to account for the several results obtained in the foregoing experiments. 



132. Conceive A B to represent .'i magnetized 

 needle, directed freely on the plane of the magnefic 

 meridian: the natural direction which this needle 

 ■would assume, were it not for the countei balance ap- 

 plied to the arm B C, would be B' A' ; and (o some 

 agent acting in the latter direction, according to our 

 supposition, it owes its directive qti.ility ; which is 

 only changed from its inclined to its horizontal posi- 

 tion by the counterweight. This agent still main- 

 taining its inclined direction, it is oh v ions, (hat if by 

 the approach of a ball, or mass of iron, it is deflected 

 out of its natural course, the needle itself will expe- 

 rience a similar change ; consequently, whatever 



may be the position of the neeille, the eireci prodiiced by the approach of a mass of 

 iron will depend entirely upon its situation with reference to the Vn\e B' A'.* 



If the centre of attraction of the mass be placed any where in the line E Q, or 

 in the plane of that line (which is supposed per|iendicular to B' A'), its action will 

 be equal upon each branch B' C, and C A', and conjcquently no motion ought to 

 ensue; and hence we see the origin and cause of the plane of no altraition. If, 

 on the other hand, the ball be elevated above the plane K Q, it will act strongest 

 upon, and attract the southern branch B' C, and conseqiicnlly the arm B C of the 

 needle will likewise ap))roach the iron. Again, when the ball is below the plane 

 E Q, then the attraction being strongest u|ion the branch of the fluid C A', it will 

 be deflected towards llic ball, and the arm C A of the needle will accordingly be 

 attracted in the same direction ; or the other ;irm B C willappear to be repelled in 

 the contrary one. All the phenoinena, therefore, attending our leading experi- 

 ments, are consistent with this view of the subject. The apparent ajiomaly 

 attending the action of the great and small hall (art. 3) is also immediately ac- 

 counted for on the same principle; for, in the one case, the centre of the ball is 

 above, and in the other below the plane of no attraction, and ought, therefore, to 

 produce opposite effects. Let us now see how far this idea will assist us in cor- 

 recting some of the anomalies observed in the preceding; suctions. 



* It is to be understood Ihat we here suppose the iron to be so far distant from 

 the needle as to be out of that particular sphere of rnfiaence, of either extrenirty, 

 adverted to in (art. 7CF). 



