BIRTH AND AFFINITIES OF CRYSTALS 129 



that the point is continually in contact with the strong solution, 

 and therefore that the crystal continues to grow rapidly in 

 this direction and slowly in others. That the liquid in con- 

 tact with a slowly growing crystal is supersaturated but only 

 metastable, I have proved by actually measuring its refractive 

 index. Again, when a rapidly growing needle approaches a 

 crystal at a part of its surface which is growing slowly, it ceases 

 to grow rapidly just before actual contact takes place — i.e. so 

 soon as it enters the zone of metastable solution surrounding 

 the latter. 



If a thin drop of solution of potassium bichromate be 

 watched as it crystallises on a warm slide, the ring of crystals 

 which form on the edge of the drop will be seen to advance 

 rapidly in a beautiful growth of needles ; the process is then 

 suddenly arrested, and the ring continues to grow quite slowly 

 with a thickening of the crystals. After a short interval the 

 sudden growth is repeated, and is then as suddenly checked ; 

 and so on until the drop is dry. The explanation is probably 

 the following: the labile solution in which the crystals first 

 appear is reduced in strength by their rapid growth, and 

 falls to the metastable condition, in which the growth is slow — 

 evaporation then raises the strength to the labile point ; rapid 

 growth again sets in, and the process is repeated. 



Other illustrations of the same recurrent process are to be 

 found (1) in " Liesegang's Rings " of silver chromate produced 

 by a solution of silver nitrate creeping through gelatine which 

 has been impregnated with potassium bichromate ; (2) in the 

 stripes of silver chloride deposited in nerve fibres by the opera- 

 tion of acid nitrate of silver acting upon the chlorides in the 

 fibre, as described by Macallum, and in capillary tubes as 

 described by Boehm. 



Another feature of crystals growing rapidly in a labile 

 solution is that they are not only apt to grow in long prisms 

 and needles, but to group themselves into the branching and 

 tree-like forms which lend such a beautiful appearance to many 

 of these growths. For example, the frost pattern on a window 

 pane, and the rapid growths of potassium bichromate just 

 described. 



This branching of needles is generally due to "twinning," the 

 process by which two crystals grow out of each other in different 

 directions at a fixed angle which depends upon the angles of 



9 



