the Laws of Magnetic Forces. 321 



variety of ways, that we can ever hope to arrive at an accurate 

 knowledge of her laws. 1 therefore submit the results which 

 I have obtained to the scientific world as matter for candid 

 consideration, having, at the same time, a proper sense of my 

 own limited powers of research *. 



PLYMOUTH, i 

 July 1. 1827. j 



* It may not be improper to state, that, in the preceding inquiries, the attracting 

 or repelling forces have been supposed to act in parallel lines. This appears to be an 

 essential condition of this species of force ; since the reciprocal influence of any two 

 points directly opposed to each other, as ab, Fig. 21, 22, 24, must be such as to 

 neutralize each other's action in relation to any other point more distant ; the action, 

 therefore, between the points immediately opposed to each other is exclusive, being 

 the nearest, and consequently the forces are parallel. 



It is, therefore, only when the attracting surfaces are of unequal extent, that it be- 

 comes necessary to take into the account any other force, as c d and ef, Fig. 22, 

 which, in a great variety of instances, are of no assignable value ; but to obviate any 

 error which can arise from this cause, it is requisite, when very powerful magnets are 

 employed, to give the attracting extremity of the bar an armature of soft iron, as re- 

 presented in Fig. 23. A, which, in diminishing from its base, terminates in a cylin- 

 drical surface exactly equal to that of the suspended body x ; by which means the an- 

 gular forces, as cd, ef, Fig. 22, are so intercepted and reduced, as to be of no as- 

 signable value. 



When the attracting surfaces are spherical and equal, it is requisite to determine a 

 fixed point in each opposed hemisphere, as x and y, Fig. 24, from which the sum of 

 all the attractions would produce the same effect as if those attractions were exerted 

 from every point of the hemispheres ; so that, in varying the distances, the intervals 

 may be estimated from these points, and not from the immediate point of contact. 

 These points I have found by numerous experiments fall within the opposed hemi- 

 spheres, at a distance equal to one-fifth of the radius of the spheres, supposing them 

 equal. 



