258 



NATURE 



YJuly 4. 1878 



FLOATING MAGNETS 



THE publication of my experiments on "Floating 

 Magnets," in the American Journal of Science and 

 in Nature, was made merely as a claim to this new 

 method of experimenting. I now send you the law of 

 the: morphology of their configurations, and show how 

 these experiments illustrate the phenomena of allotropy, 



GH 



isomerism, expansion or solidification of water, bismuth 

 antimony, &c., the atomic hypothesis, and the kinetic 

 theory of gases. 



The configurations of the floating magnets given in 

 this paper are reduced to half-size. They were obtained 

 as follows : — A cylindrical magnet, 387 millimetres long 



-9 



and thirteen millimetres in diameter, was clamped with 

 its lower end sixty millimetres above the plane in which 

 were the ends of the floating magnets. 



After each configuration was formed the tips of the 

 needles were dotted with printer' s ink, and a flat piece 

 of cardboard was carefully lowered on to the configura- 



tion, which was thus printed on the card. The points 

 formed in this way were placed on drawing-paper, and 

 the imprinted points were pierced with a needle. Thus 

 the centres of the magnets were located, and around 

 these points were drawn the circles of the element of 

 the configurations. The configurations here given are 



one-half the size of the prints taken from nature, "with 

 all their imperfections on their heads," produced by the 

 unavoidable unequal magnetization of the component 

 needles. 



These configurations are numbered from 2 up to iB^; 

 the numbers indicating the numbers of floating magnets 



in the configurations. Where a, b, and c occur under a 

 configuration they show the order of their stability. Thus 

 5 « i s more stable than 5 b and 6 a than 6 b. 



Th e law of the morphology of these forms is as 

 folio ws : — They are divided into primaries, secondaries, 

 tertiaries, &c. The primary configurations are from 2 up 

 to 9«. 



The secondaries begin with 9 (one might even say with 

 b and c oi 8). These secondary configurations have the 



stable primaries for nuclei. Thus configurations 9, 10, 1 1, 

 12, 13, 14, 15, 16, 17, i8rtr, and 1 8<5, have respectively 2, 

 3, 3, 4, 4, 5 (flattened), da (which- is "5 flattened" ex- 

 panded to a regular pentagon), '], pointed {compressed f) 

 towards a vertex of the hexagon, 7, 7, 8. 



Nineteen needles form the first configuration of the 

 tertiaries. This is formed of 9 as nucleus, surrounded 

 by 10 floating magnets. 



Twenty has 9 for nucleus, with 1 1 circumposed ; but 

 this form is unstable, and soon changes into Fig. 20, 

 which has 10 magnets for nucleus with 10 circumposed. 

 This is the only instance (except the flattened pentagon, 

 Fig. 14) I have found where a nucleus is changed in form 

 by the action of the circumposed magnets. This nucleus 

 of 20 cannot be formed without the circumposed magnets, 

 as in Fig. 20, 



Twenty-two has 11 for nucleus, surrounded with 11 

 magnets. 



Twenty-three h?is 11 for nucleus, with 12 circumposed 

 needles, arranged parallel to nucleus. 



Twenty-four is formed of 1 1 for nucleus, surrounded 

 with 1 2, and otie opposite the base of .'. 



Twenty-five IS iormed of 13 for nucleus, with 12 cir- 

 cumposed, and parallel to nucleus. 



