18 INTRODUCTORY OBSERVATIONS. 



thickness remaining unchanged, and have thence deduced for- 

 mulae belonging to an indefinitely extended plate of uniform 

 thickness. From these it follows, that when the point p, and 

 the centres of the magnetic forces are situate on opposite sides of 

 a soft iron plate of great extent, the total action on p will have 

 the same direction as the resultant of all the forces, which would 

 be exerted on the points p, p ', p",p", etc. in infinitum if no plate 

 were interposed, and will be equal to this resultant multiplied 

 by a very small constant quantity : the points p, p ', p", p'", &c. 

 being all on a right line perpendicular to the flat surfaces of the 

 plate, and receding from it so, that the distance between any two 

 consecutive points may be equal to twice the plate's thickness. 



What has just been advanced will be sensibly correct, on 

 the supposition of the distances between the point p and the 

 magnetic centres not being very great, compared with the plate's 

 thickness, for, when these distances are exceedingly great, the 

 interposition of the plate will make no sensible alteration in the 

 force with which p is solicited. 



When an elongated body, as a steel wire for instance, has, 

 under the influence of powerful magnets, received a greater 

 degree of magnetism than it can retain alone, and is afterwards 

 left to itself, it is said to be magnetized to saturation. Now if 

 in this state we consider any one of its conducting elements, the 

 force with which a particle p of magnetism situate within the 

 element tends to move, will evidently be precisely equal to its 

 coercive force f, and in equilibrium with it. Supposing there- 

 fore this force to be the same for every element, it is clear that 

 the degree of magnetism retained by the wire in a state of satu- 

 ration, is, on account of its elongated form, exactly the same as 

 would be induced by the action of a constant force, equal to/, 

 directed along lines parallel to its axis, if all the elements were 

 perfect conductors; and consequently, may readily be deter- 

 mined by the general theory. The number and accuracy of 

 COULOMB'S experiments on cylindric wires magnetized to satu- 

 ration, rendered an application of theory to this particular case 

 very desirable, in order to compare it with experience. We have 

 therefore effected this in the last article, and the result of the 

 comparison is of the most satisfactory kind. 



