SURFACES OF DISCONTINUITY 673 



The gist of the long footnote on page 320 is that since two 

 pieces of ice, for example, do not freeze together spontaneously 

 but only under pressure, the free energy of the discontinuous 

 region formed between the two pieces on freezing, denoted 

 by (T// is not less than, and is most probably greater than, the 

 sum of the free energies of the two surfaces in existence before 

 the regelation, denoted by 2(tjw. 



The argument concerning crystalline solids follows the same 

 course. To enable the reader to grasp the reason for the second 

 part of the expression on page 320, Figure 11 is supplied. It 

 represents a section of the crystal at the edge V which is sup- 

 posed to extend at right angles to the plane of the paper; BE 

 is part of the section of the surface s by the paper, AB a. part of 



Fig. 11 



the section of s'; CF is a part of the section of the surface s after 

 growth of the crystal, so that the angle EBC is w', and CD is 

 equal to bN. The face s' has, as far as the phenomena around 

 the edge at D are concerned, increased by an area I'BC, i.e. 

 V • CD cosec co' or V • cosec co' 8N; the face s has decreased by an 

 area I' ■ BD or V cot w' 8N. Of course if co' is greater than a right 

 angle, at any edge, the term involving cot co' in the correspond- 

 ing portion of the summed expression will be essentially nega- 

 tive and the term will be virtually an addition term, as is clear 

 from the fact that at such an edge s increases in area. 



The argument on page 322 concerning stability follows 

 precisely the same course as those employed earlier in the case 

 of fluids, on which we have already commented fully. It should 

 offer no difficulty. Nor is there anything in the three following 



