Royal Society, 149 



towards each other no action whatever, either of attraction or re- 

 pulsion. It has long been known that the magnetism of iron is im- 

 paired by heat; and it has been generally believed that a certain 

 degree of heat destroys it entirely. The author finds, however, that 

 this opinion is not correct ; for he shows that, by applying more 

 powerful tests than those which had been formerly confided in, 

 iron, nickel and cobalt, however high their temperature may be 

 raised, still retain a certain amount of magnetic power, of the same 

 character as that which they ordinarily possess. From the different 

 temperatures at which the magnetic metals appear to lose their pe- 

 culiar power, it had formerly been surmised by the author that all the 

 metals would probably be found to possess the same character of mag- 

 netism, if their temperature could be lowered sufficiently ; but the re- 

 sults of the present investigation have convinced him that this is not 

 the case, for bismuth, tin, &c. are in a condition very different from 

 that of heated iron, nickel or cobalt. 



The magnetic phenomena presented by copper and a few other 

 metals are of a peculiar character, differing exceedingly from those 

 exhibited by either iron or bismuth, in consequence of their being 

 complicated with other agencies, arising from the gradual acquisi- 

 tion and loss of magnetic power by the iron core of the electro- 

 magnet, the great conducting power of copper for electric currents, 

 and its susceptibility of being acted upon by induced currents of 

 magneto-electricity, as described by the author in the first and se- 

 cond series of these researches. The resulting phenomena are to 

 all appearance exceedingly singular and anomalous, and would seem 

 to be explicable only on the principles referred to by the author. 



Pursuing his inductive inquiries with a view to discover the pri- 

 mary law of magnetic action from which the general phenomena 

 result, the author noticed the modifications produced by different 

 forms given to the bodies subjected to experiment. In order that 

 these bodies may set either axially or equatorially, it is necessary 

 that their section, with reference to the plane of revolution, be of an 

 elongated shape : when in the form of a cube or sphere they have 

 no disposition to turn in any direction : but the whole mass, if mag- 

 netic, is attracted towards either magnetic pole ; if diamagnetic, is 

 repelled from them. Substances divided into minute fragments, or 

 reduced to a fine powder, obey the same law as the aggregate masses, 

 moving in lines, which may be termed diamagnetic curves, in con- 

 tradistinction to the ordinary magnetic curves, which they every- 

 where intersect at right angles. These movements may be beauti- 

 fully seen by sprinkling bismuth in very fine powder on paper, and 

 tapping on the paper while subjected to the action of a magnet. 



The whole of these facts, when carefully considered, are resolva- 

 ble, by induction, into the general and simple law, that while every 

 particle of a magnetic body is attracted, every particle of a diamag- 

 netic body is repelled, by either pole of a magnet. These forces 

 continue to be exerted as long as the magnetic power is sustained, 

 and immediately cease on the cessation of that power. Thus do 

 these two modes of action stand in the same general antithetical re- 



