510 Professor Thomson [May 14, 



action of different gases in determining tbe crystallisation. These 

 ideas were followed by the statement of Ziz, that not only air but 

 also solids were capable of acting as nuclei when dry, but that 

 when wet, or boiled with the solution, or placed in it when hot 

 and allowed to cool at the same time with it, they lost their effect. 

 The activity of air, however, as in itself a nucleus, has been sub- 

 sequently shown by Lowel to be incorrect ; but at the same time 

 he admits that solid bodies after exposure to air become active. 



In 1851 two chemists, Selmi and Goskynski, introduced the 

 explanation that dry air is active by virtue of its getting rid of the 

 water at the surface, thus producing small crystals which continue 

 the action. This seems to be a repetition of the theory propounded 

 by Gay Lussac, who held that the air absorbed at the surface 

 precipitated a portion of the salt in the same way that one salt 

 precipitates another, and that this deposition continued the crystal- 

 lising action. By an elaborate series of investigations, conducted 

 in 1866, and also during later years, the French savants, MM. 

 Gernez and Violette, came to the conclusion that there is only 

 one nucleus for a supersaturated solution, and that is a crystal 

 of the body itself; also indicating in certain experiments that 

 substances which possess the same crystalline form and chemical 

 structure may be found active to supersaturated solutions of each 

 other. They have conclusively shown at the same time that 

 heated or washed air, or bodies of different constitution but chemi- 

 cally clean, remain perfectly inactive as regards their supersaturated 

 solutions. 



That disruption of the solution may take place without causing 

 crystallisation when the body added is perfectly clean may be 

 easily shown by carefully removing the cotton plug from a solution 

 of sodium sulphate containing only two-thirds its weight of crystals. 

 On introducing a perfectly clean platinum wire no crystallisation is 

 induced, but on removing the wire and touching it with a substance 

 which has not previously been rendered chemically clean, and on 

 again introducing it in the flask, the moment it touches the liquid 

 crystallisation is at once induced. 



This question of the nuclear action of substances upon these 

 solutions has received considerable attention from English chemists, 

 most notably from Mr. Tomlinson, Professor Liversidge, and 

 Professor Grenfell, all of whom have conducted elaborate researches 

 on the subject, leading, however, to different conclusions with each 

 investigator. The arguments held by Tomlinson are strongly in 

 suj)port of some physical cause for the phenomenon, and that the 

 result may be brought about by the action of substances possessing 

 no chemical relations to the solutions experimented on. Professor 

 Liversidge, on the other hand, has tried the action of several sub- 

 stances with the most careful precautions, that their addition to the 

 solutions should be effected without contamination from, or access of, 

 the external air to the flask during the experiments. The result of 



