326 Mr. Barlow's Experiments on 



several arrows the dii'ection of rotation as excited by the mag- 

 net and lamp in these four cases. Then it is obvious that in 

 fig. 11, answering to the case of the lamp applied at E, the 

 tendencies to rotation are all in one direction, and we ought 

 therefore to expect, as is actually the case, the rotation to be 

 very rapid. In figure 12, two of the forces are in one direc- 

 tion, and two in the other ; and therefore if these forces were 

 all equal, we ought to have no motion. But the force at D is 

 very considerable in comparison with that at F or G,both on 

 account of the immediate application of the lamp at E, and 

 the division of the circuit into three branches at O ; and the 

 direction of the two latter forces on the respective levers, 

 which being oblique there are necessarily less effective in 

 producing rotation. Again, the force at E is also considera- 

 ble in consequence of the proximity of the magnet ; so that the 

 superior forces at D and E, conspiring in direction, will ovei'- 

 power the other two weak and oblique forces, and produce a 

 very considerable rotation, although inferior to the former. 



In fig. 13, the forces at E and G will be the superior forces; 

 and as they conspire in direction, they will overpower the 

 two inferior forces at D and F, which are opposed to them, 

 and a considerable rotation will again ensue. 



In fig. 14, the two superior forces also conspire, but with 

 this peculiarity, that the moment the motion ensues, and the 

 arm F arrives at F' and E at E', the direction of these two 

 forces no longer assists in giving rotation, being then both in 

 the direction of the radii from the centre, and the resultant 

 acts to bring the machine directly towards the magnet, and 

 thereby to convert the rotation into direction, which the ex- 

 periment strongly exhibits. 



When the south pole of the magnet is opposed to the wire, 

 all these directions of motion will be reversed, and then of 

 course the point of neutralization will be at G, fig. 1 fj : which 

 explains the apparent anomaly of the point of no action being 

 to the right hand of the north pole and to the left of the south 

 pole. In fig. 13 we have seen that with the lamp applied at 

 G the motion is to the left ; and when at E, as in fig. 1 1 , it is 

 to the right ; it follows, therefore, that between these two 

 there ought to be some position of equilibrium as G', fig. 13, 

 and where no motion of course ought to ensue. But in this 

 case, instead of the forces at G' and E' being directed yJow/ 

 the centre as in the case of fig. 14, they are directed to the 

 centi'e. So that this state of equilibrium differs from the for- 

 mer in this, — that in the former, the equilibrium is one of 

 stability, and in the latter of instability, and consequently 

 very difficult to exhibit: in fact, the slightest inclination of 



the 



