268 THE ROYAL SOCIETY OF CANADA 



For: Sodium thiosulphate 10^^ 



Sodium chlorate lO^lO'i" 



Potassium alum 10"^° 



Rochelle salts 10"ii 



Borax lO'i" 



Salol 10"8-10"w 



Particles of a mass of 10'^^ are easily microscopically visible; 

 but Ostwald expresses a view that the limiting size to induce crystal- 

 lization is about lO'^'^ 



From both the theoretical and experimental sides this con- 

 clusion may be combated. There is no a priori reason why a perfect 

 crystal, however small, should not bring about crystallization. 



X-ray spectroscopy has shown that crystals are comparatively 

 simple, and that a perfect crystal may be formed from only a few 

 molecules. Barlow's views agree with this, and Professor Graham 

 has informed me that he sees no reason why a perfect crystal of 

 hydrated sodium sulphate should not be formed from four molecules. 

 If this be so, the crystal particle bringing about precipitation need 

 not be larger than 10'^°— 10^^ gram. 



It has been noticed that sodium sulphate decahydrate, which has 

 effloresced, is as active in bringing about crystallization in super- 

 saturated solutions as the perfect salt. Glass hairs, drawn lightly 

 over sodium sulphate crystals, are sufficient to bring about crystalliz- 

 ation, although the adhering matter probably weighs not more than 

 10'^ gram. These glass hairs with sodium sulphate, rochelle salts, 

 sodium acetate, etc., have been kept in tubes, dried by phosphoric 

 anhydride at a pressure of 1/1000 millimeter of mercury, for over a 

 month and were still active. Two cases, ammonium acetate and salol, 

 yielded negative results; as their vapour pressures were high the 

 materials had quite disappeared. If we accept Morley's view that the 

 vapour pressure of water is reduced to 1/40,000,000 of an atmosphere 

 in the presence of phosphoric anhydride, one can calculate that the 

 mass of the decahydrate present is not greater than lO"''*. 



DeCoppet regards these dried particles as anhydrous crystals 

 isomorphous with the original crystal. This appears unlikely, since 

 it postulates at least three anhydrous forms; the true anhydrous, 

 stable above 33°C., the anhydrous heptahydrate, and the anhydrous 

 decahydrate. The fact that the sterilizing temperature for the dried 

 decahydrate is exactly 33° indicates, if it does not prove, the presence 

 of decahydrate crystals. 



As it is difficult to avoid contamination in removing the glass 

 hairs from the drying chambers, this question was tested in another 



