90 



inverse sesquiduplicate ratio, or f power of the distances ; between 

 ^rd and fths as the squares of the distances inversely. From the 

 ^th to -i- of the limit of action it may be as the f power of the 

 distance inversely ; within less than ith, it will be generally as the 

 simple distance inversely. 



On a further review of these laws of magnetism, it is evident that 

 the immediate effect of an increase or decrease of distance, is an in- 

 crease or decrease of the effective magnetism on which the total or 

 reciprocal force depends. Thus taking the cases just quoted, it will 

 be seen that the total force is always as the square of the induction, 

 whatever be the resulting law of the attraction. Hence the force 

 may as well be taken as the square of the quantity of effective 

 magnetism directly, as some power of the distance inversely. 



The author admits the difficulty in the way of the employment of 

 such terms as quantity of magnetism, magnetic charge, and the like, 

 and therefore only employs them according to the common accepta- 

 tion of such terms, and not as referring to any particular hypothesis : 

 he thinks there must necessarily be in such inquiries an element fairly 

 enough expressed by the general term quantity as expressive of the re- 

 lative or absolute magnitude of the cause, whatever it be, upon which 

 the observed effects depend, and thinks it so far essential to obtain 

 exact quantitative measures. In electricity we may estimate the 

 charge conveyed into a battery by means of the unit measure, and we 

 can at pleasure operate v/ith one-half, one-third, &c. the quantity of 

 electricity numerically expressed ; but we have as yet no such mea- 

 sure in magnetism, and we are quite uncertain as to the quantity of 

 effective magnetism in operation. The author hence endeavours to 

 verify the law of magnetic charge just mentioned by a direct quan- 

 titative experimental process. A cylindrical rod of soft iron being 

 gurrounded by three successive coils of covered copper wire, was 

 placed under the trial cylinder of the magnetometer and exposed to 

 the operation of one or more precisely equal and similar batteries ; 

 one coil being appropriated to each battery. It is inferred that if 

 one battery and one coil produced one measure of magnetism, two 

 batteries and two coils would develope two measures, and so on ; so 

 that we should have only to determine the attractive force under 

 this condition ; now the attractive forces were found to be as the 

 square of the number of batteries in action upon this cylinder, that 

 is to say, as the square of the magnetism induced in the iron ; hence 

 the quantity of magnetism is as the square roots of the reciprocal 

 forces. If therefore the reciprocal force between a magnet A and 

 a cylinder of soft iron taken at a constant distance were represented 

 by an equivalent of 4 grains, whilst the similar force with a magnet 

 B at the same distance were represented by 9 grains, then the effect- 

 ive quantities of magnetism and operation in each case would be as 

 : ^9, that is as 2 : 3. 



Availing himself of this law, the author endeavours to deduce ex- 

 perimentally the magnetic development in different points of a regu- 

 larly tempered and magnetized bar, taken between the magnetic 

 centre and extremities ; and he finds by a very careful manipulation, 



