7idy \, 1878] 



NATURE 



259 



Twenty-six is formed of 14 for nucleus, with 12 magnets 

 circumposed. 



Twenty-seven is formed of 15 for nucleus, with 13 

 magnets circumposed. 



Twenty-eight is formed of 14 for nucleus with 13 

 circumposed. 



Twenty-nine is formed of 16 for nucleus, with 13 cir- 

 cumscribed. 



Thirty is formed of 17 for nucleus, with 13 circum- 

 scribed. 



Thirty-one is formed of 18 for nucleus, with 13 circum- 

 scribed. 



Thirty-two begins the Quaternary Configurations, for 

 it is formed of 19, with 13 circumposed magnets. 



Thirty-three is formed of 20 for nucleus, with 13 

 magnets circumposed. 



Thirty-four is formed of 21 for nucleus, with 13 mag- 

 nets circumposed. 



Thirty-five is formed of 22 feet for nucleus, with 13 

 magnets circumposed. 



Thirty-six is formed of 23 for nucleus, with 13 mag- 

 nets circumposed. 



Thirty-seven is formed of 24 for nucleus, with 1 3 cir- 

 cumscribed magnets. 



Thirty-eight is formed of 26 with 12 magnets cir- 

 cumscribed. 



The expansion of liquids, like ice and antimony, on 

 solidification, allotropy and isomerism, are illustrated 

 by the fact that different configurations formed of the 

 same number of atoms hare densities inversely as their 

 areas. 



Thus lb is about i^th greater in area than 5^:. So if 

 5a represent water at 0° C., ^b may stand for ice at 0° C. 



Similarly in allotropy if ta stand for graphite, then db 

 may stand for diamond, and the three forms of titanic 

 oxide, rutile, brookite, and anatase, and their different 

 densities may be illustrated by configurations \a, Sb, Sc. 



tSa 



If 6b stand for calcite, then 6a will stand for its isomer 

 anagonite. 



The law ruling the density of the configuration is evi- 

 dently that a central magnet always expands the contours 

 of the configuration. For example, compare Figs. 4 

 and 5^, 5a and 6a, 6b and 6a, 14 and 15. 



These configurations— at least the stable ones — can be 

 obtained by suspending the magnets by fine silk fibres. 

 I have thus obtained all the stable forms ; and the plan 

 proposed to me by my friend Prof. Rood will no doubt 

 give these configurations. He proposed to me to suspend 



