Lifting Power of the Electro-Magnet. 503 



If a second spiral or helix be placed symmetrical to the first at a, the effect 

 at A is the resultant of four components, the two first named, and the induc- 

 tions of the branch aBA = z + ^', and of aDC A =c — k — z. The same principle 

 applies to any additional spires, and if they be symmetrically placed on the 

 branches AC and BD, the forces developed at A and B are equal. (c) 



If the forces excited by each spire were transmitted undiminished by in- 

 duction, and if they were compounded by mere addition, it would be an easy 

 matter to compute the force at either of these points ; but it will subsequently 

 be shown that these resultants are always less than the sum of the com- 

 ponents, and that the law of their composition, as well as that of the transmis- 

 sion by induction, are yet unknown. (c?) 



Now let the magnet be discontinuous at A and B: AB becomes a keeper ; 

 but this makes no change, except as far as the separations there make the in- 

 ductive propagation through the circuit more difiicult. However closely the 

 surfaces of contact may be fitted, and though they are pressed together very 

 powerfully by the attraction of the magnet, it is certain that they are separated 

 by an interval far greater than the distance between the molecules of iron; 

 it is even possible that this interval may be sufiicient to change the law of 

 inductive propagation. (e) 



The following experiments will illustrate the statements (a) and (c) : — 



The part CD of the magnet (described in my second paper, vol. xxii., page 

 508) is a moveable slide, which can be set to give AC any value from to II'. 

 If it be removed, and both arms be excited, the force at A is only the resultant 

 of the inductions through z and z-\-k^ themselves diminished by the interruption 

 of the circuit. If, also, for AB be substituted a block of brass, to which cylin- 

 ders of iron are fixed 2' high, and of the same diameter (that of the magnet), 

 then, with the slide in its place, we have the resultant of induction through (5;), 

 and that through (c — k — zy, if it be removed, there is left only (2^). The 

 helices used are described in No. 11, page 519, and K& — k is 6 inches. 



The temperature is given to show that no sensible portion of the differences 

 is due to a variation of it ; but in these, as in most of the other results given in 

 this paper, no temperature correction is made, as it must be specially investi- 

 gated for each altered arrangement of the magnet. 



3u2 



