286 Mr HARRIS'S Experimental Inquiries concerning 



was placed a magnet m, nine inches long, and of the same breadth 

 and depth as the iron above ; the whole was then transferred 

 under the suspended cylinder a?, as in Fig. 5, it being previously 

 ascertained that the magnet m might be alone approximated 

 within two inches of the iron #, without any sensible effect being 

 produced on the index. In this arrangement, therefore, the in- 

 dex could not become influenced, except by the magnetic de- 

 velopement induced in the intermediate substance be; and thus, 

 by varying the distance cd, and at the same time preserving the 

 distance ab, by means of the screw at S, constant, it was easy to 

 determine the law according to which the magnetic developement 

 in the iron proceeded ; the force of the magnet m being consi- 

 dered a constant quantity, but its distance from the iron a va- 

 riable one. 



16. For it will be readily admitted, that any polarity which 

 the attracting masses of iron be and x could be supposed to ac- 

 quire by position might be considered as invariable and fixed 

 throughout the experiment, and therefore could not affect the 

 result, and must be otherwise a quantity so small in relation to 

 the means by which the other forces were made sensible, that it 

 could not have any assignable value, as the masses of iron x and 

 be would not alone evince any attractive force, so as to be sen- 

 sible by the index, however near they were approximated. 



17- For similar reasons, the operation of the distant polarities, 

 as they became developed in the attracting masses of iron # and 

 be, could not be supposed to exert any sensible influence in com- 

 plicating the result, as will also appear by considering the cir- 

 cumstances under which these polarities are placed. Thus, when 

 two magnets A and B, Figs. 6, 7. &. 8. are opposed to each other at 

 their dissimilar poles, then, in a purely theoretical sense, and ac- 

 cording to the most evident of magnetic experiments, N attracts 



