126 THE FORMS OF WATER IN 



326. Yon have not forgotten onr conversation re- 

 garding ' atomic poles ' (§ 10), and how the notion 

 of polar force came to be applied to "crystals. With 

 this fresh in yonr memory, you will have no great diffi- 

 culty in understanding how expansion of volume may 

 accompany the act of crystallisation. 



327. I place a number of magnets before you. They, 

 as matter, are affected by gravity, and, if perfectly free, 

 they would move towards each other in obedience to 

 the attraction of gravity. 



328. But they are not only matter, but magnetic 

 matter. They not only act upon each other by the 

 simple force of gravity, but by the polar force of 

 magnetism. Imagine them placed at a distance from 

 each other, and perfectly free to move. Gravity first 

 makes itself felt and draws them together. For a time 

 the magnetic force issuing from the poles is insensible ; 

 but when a certain nearness is attained, the polar force 

 comes into play. The mutually attracting points close 

 up, the mutually repellent points retreat, and it is easy 

 to see that this action may produce an arrangement of 

 the magnets which requires more room. Suppose them 

 surrounded by a box which exactly encloses them at 

 the moment the polar force first comes into play. It 

 is easy to see that in arranging themselves subsequently 

 the repelled corners and ends of the magnets may be 

 caused to press against the sides of the box, and eveu 

 to burst it, if the forces be sufficiently strong. 



