Section III, 1919 [265] Trans. R.S.C. 



The Crystallization of Supersaturated Solutions and Super- 

 Cooled Liquids 



By D. McIntosh, F.R.S.C. 



(Read May Meeting, 1919.) 



Few natural phenomena show so much beauty or appear so 

 wonderful as the formation of crystals. From a homogeneous solu- 

 tion, the molecules, in an instant, rush into position and form well 

 defined bodies of various sizes and figures. It is not strange then, that 

 for many years the formation of crystals has been studied, and that 

 amongst other problems, chemists and physicists have attempted to 

 define the conditions and explain the causes of crystallization from 

 supersaturated and supercooled liquids. To give any extended 

 account of this work is unnecessary; those interested will find an 

 excellent account in Ostwald's Lehrbuch. 



Many molten substances may be cooled far below their melting- 

 points without solid matter appearing; and many solutions may be 

 made which remain liquid although they hold in solution far more of 

 the solute than would dissolve at that particular temperature. Mechani- 

 cal shocks, treatm.ent with foreign substances, changes of pressure, 

 of temperature within certain limits, produce no crystallization. But 

 precipitation is induced by the smallest crystal of the substance in 

 solution. And this fact has been made the basis for a "germ" theory, 

 which has been extended by Ostwald. 



Ostwald recognises two fields in his supersaturated or super- 

 cooled system — a labile and a metastable. In the former, crystalliza- 

 tion may occur through mechanical shocks, etc.; in the latter, only 

 by the action of a crystal of the material in solution, or isomorphous 

 with it. For instance, in a supersaturated solution of sodium, sul- 

 phate crystallization of the heptahydrate may be brought about by 

 mechanical means at 10°C. The decahydrate, however, can be pre- 

 cipitated only by the decahydrate crystals. On cooling the solution 

 to — 5°C., the decahydrate passes from the metastable to the labile 

 field, and at that temperature can be precipitated by mechanical 

 means. 



This theory, accepted by chemists for a decade, was proven 

 untenable by Young. For a great mechanical shock can be shown 

 to be more effective in inducing crystallization than a weak one; 

 the greater the shock the more the metastable field is narrowed. 



Sec. Ill, Sig. 17 



