272 THE ROYAL SOCIETY OF CANADA 



salol was inoculated in the manner described. Although the solution 

 was absolutely clear it produced crystallization invariably. Indeed, 

 I have had a solution containing such fine particles that they had not 

 fallen 5 centimeters after 40 days; but as my solutions were not in a 

 constant temperature bath, this can in part be ascribed to temperature 

 variation. 



The results from the microscopic measurements were: after 24 

 hours, diam. 2v, mass 4X10'^^; after 96 hours, diam. 0-5y, masë 

 6X10'^^; final results, diam. 0-2y, mass 4X10"^^, density of salol, 

 1-27. 



While these results are only a first approximation, I have every 

 confidence that the necessary mass to cause crystallization does not 

 exceed this amount. Perrin, in his determination of the value of N, 

 shows that a gamboge solution, (density 1-2) containing particles of 

 a diameter of -42 v comes to a stable state in a few hours, and that 

 the number of the particles decreases to one-half for each 30 u ascent 

 in the liquid. In one millimeter the number of particles would dim- 

 inish to ^^^, or become vanishingly small. This leads me to believe 

 that the particles were even smaller than I had estimated. 



Finally, it may be noted that salts which readily give super- 

 saturated solution with water fall into two principal classes: those 

 containing water of crystallization, and organic substances. In both 

 classes destruction of the crystal form is brought about by moderate 

 heating, so that the solutions are easily "sterilized." 



The anhydrous salts forming supersaturated solutions are mainly 

 compounds such as the nitrates and chlorates which are easily des- 

 troyed by heat and whose crystals are therefore not likely to be present 

 on the surfaces of the containing vessels. 



In the case of anhydrous salts, the failure to form supersaturated 

 solutions easily can be accounted for by the separation of the solid 

 from drops on the sides of the vessels containing the solutions. These 

 minute crystals fall into the solution and induce crystallization. 



