508 The Rev. R. T. Robinson's Experimental Researches o?i the 



There is the same decrease of L, as c and s increase ; but it is much slower, 

 as is especially seen by comparing the two first sets of ratios with the fifth and 

 sixth columns, which give the ratios of i* and L, obtained with the spirals for 

 A = 2'l, with their other values. This fact arises apparently from the more 

 powerful action of the spiral in its plane, being neutralized by its feeble action 

 out of that plane, and from the greater distance through which it must act by 

 induction. But in any case it is obvious, that the lifting power of an electro- 

 mao^net is increased by making the length of the circuit as short as possible, and 

 by bringing the wire of the helices, or spirals, as near the poles as is compati- 

 ble with having the necessary number of spires, as was stated in my second 

 paper, p. 522. It will be shown, however, that another plan of arrangement 

 should be adopted when the magnet is intended to act at a distance. 



I have tried in vain to find any simple law connecting L with c and z, and 

 it probably is not to be found by mere experiment, for — 



1. The direct action of a spire on that section of the magnet which is in its 

 plane, can be computed by my equation (4) ; and that on any other by equa- 

 tion (10). Each of these actions is transferred to the sections A and B by 

 induction ; but the law of this induction is not known, and it does not seem pro- 

 bable that it will be possible to evolve it from an effect so complex. Yet 

 more, these inductive actions influence the power of the current itself to ex- 

 cite the intermediate sections in a manner which is still undetermined. 



2. In these experiments the observed force is, as I have stated in (d), the 

 resultant of four components ; and we do not know the law of this composi- 

 tion. Each of them is the result of an induction through a certain length of 

 iron, but we know nothing of their relative magnitudes. We may get a gene- 

 ral notion of the composition of magnetic force by forming a table, such as Table 

 xn., and comparing its L and t^. I think it cannot be doubted that if -f pro- 

 duces L, and •^' L' ; ir + ^lr' will produce the resultant of L and L'. That this 

 may be more easily examined, I have reduced the L to those corresponding to 

 an arithmetical series of -f, for this magnet, which is 2' diameter with c- 16'-2 ; 

 and have done the same for the hoUow magnet described in my first paper, 

 whose external diameter = 2', and its c = 36 ; for a small one, to be described 

 hereafter, of diameter = 1', and c = 27'-7 ; and for one 3' diameter, and c= 36'-0. 



1 



