200 Mr. J. J. Waterston on some Electrical 



librium among the lines requires that they be pressed as much 

 inwards towards q r as outwards ; but tracing round the sphe- 

 rical surface till we come to the extreme outward points opposite 

 q and r, it is obvious that, although the excited roots of these 

 points may have others adjacent mutually repellent on all sides, 

 yet the lines that issue from these extreme points must in 

 their curvilinear course Im, uv take directions that cannot have 

 juxtaposition except at the issuing points on the surface. What, 

 then, prevents them yielding to the repellent forces of the inte- 

 rior adjacent lines ? Let m (fig. 38) represent an ultimate me- 

 tallic chemical atom in an electric line P N. If such a condition 

 could be realized, there is every reason to believe that the side 

 towards N would be positively electrified, and the side towards P 

 negatively electrified, as represented by the V symbols, and that 

 it is thus being pulled equally from both sides with intensity in 

 accordance with the nearness of adjacent lines. Now let us sup- 

 pose a similar atom situated at each symbol thus (fig. 39). We 

 have, instead of an electric line, a line of electrified molecules 

 each of which is drawn by an equal force in opposite directions, 

 giving to the whole line a longitudinal contractile tension. Now 

 supposing such a line to be curved, this contractile force will 

 engender a lateral compressing power towards the concave side ; 

 and as all the disengaged electricity in this system consists of 

 lines similarly curved, i. e. with similar convexity one over the 

 other, the lateral packing power is cumulative } and the central 

 engaged lines will be packed together by means of the curvili- 

 nearly derived power of the whole of the system exterior. 



The proof of the existence of this contractile power in a line is 

 obtained by employing the method of concentric films as in 52. 

 Suppose B and all the concentric films to contract simulta- 

 neously and concentrically upon A by the electric action between 

 the adjacent surfaces in degrees respectively proportionate to 

 their distance from it. Here is a system of lines expending 

 their integral, and in doing so exerting a contractile force 

 throughout, which must therefore have had a contractile tension 

 as potential antecedent, a force acting at each point of a line 

 towards its opposite extremities. A line composed of india-rubber 

 has a contractile force which might be thus defined ; but it differs 

 from an electric line in this, that in the rubber line the force is 

 the same at every point ; but in the electric line the force depends 

 on the distance of adjacent lines, and is thus variously distri- 

 buted according to circumstances. Thus, e. g. 3 in the concen- 

 tric spherical arrangement of 52. the contractile tension dimi- 

 nishes outwards regularly. When the electric lines are closely 

 engaged, the contractile tension must be nearly uniform. In 

 the systems figs. 34, 35 it may be small at the root, gradually 





