November 4, 1920] 



NATURE 



l^i 



From a larjje number of observations the 

 authors have drawn the following conclusions : 

 (i) Crvstal growth always took place by gradual 



i.e. crystal growth does not depend on the rela- 

 tive size of crystals ; (3) the relative orientations 

 of the crystal which is being grown into and that 



Flo. >— and heat. 



K.... 4. -,r,l ll.al. 



V ■'.. «[. 4ih lieiil. 



MnfnifkAiion loodUm*. 



boundary migration, and not by coalescence; 

 i2\ growth may take place either of a large crystal 



small one, or of a small into a large — 



'. 2662, VOL. 106] 



Fir.. 6.— Sami <• Ki«. 5, reuoli'htd »n.I r«-«lch*d. 

 R«duc«d by ont.ttitth. 



which is growing do not apix-ar to affect growth ; 



(4) a crystal which is itself Ijeing invaded by one 



" crystal may grow at the same time at the expense 



