510 The Eev. T. R. Robinson's Experimental Researches on the 



raised by a force Z, applied at its middle point at right angles to the plane of 

 these surfaces, we have 



L = 2M; 



M being the force at each surface. If, however, one arm only be excited, the 

 polar forces are unequal ; the weaker gives way first, and the keeper rises, 

 turning round the outer edge of the excited arm. In this case we have, call- 

 ing r the radius of the arm, 



Ur^L'--r^ X {M-M). 

 ^k, -\-r 



A second equation is obtained by assisting M' with a downward pressure P, 

 applied exactly in the axis of the imexcited arm, and sufBcient to make it pre- 

 vail ; then M gives way, and thence / 



from which M and M' are found. 



In this magnet r= 1, ^"= 6, whence 



To apply the pressure P a lever was used, whose ratio = 5-8, turning on a cen- 

 tre in a strong vertical bar of iron screwed to the slide ; weights of an / re- 

 quired amount were hung at its extremity, and it pressed on a rod of steel, 

 pointed at each end, one of which was set in a hole punched in its lower -ur- 

 face, the other in the centre of the screw which is fixed in the keeper, with :ts 

 head entering the central pole in the magnet to serve as a guide. For each 

 position of the slide one of these rods was provided, so adjusted that the lever 

 was horizontal, and of course the rod exactly vertical. The bar prevented the 

 application of the larger helices to the arm BD, near which it was fixed, so 

 that the force M belongs to A, M' to B. 



The first set were taken with the spirals A, A'; -^ is nearly 124-69, and the 

 resulting M and M' are reduced to that by interpolation. Only two are avail- 



