BIRTH AND AFFINITIES OF CRYSTALS 123 



at a higher temperature than that corresponding to the appear- 

 ance of the second shower in an open vessel. For example, a 

 48*2 per cent, solution of sodium nitrate is saturated at 26 , and 

 yields the dense shower at 16 ; a 51 per cent, solution is saturated 

 at 39 , and yields the dense shower at 30 ; and these solutions 

 enclosed in sealed tubes only yield a shower on shaking at 16 

 and 30 respectively. 



During- the visit of the British Association to South Africa 





£ 







50 

 Cnjicrnf ralrriH.. 



Fig. 1. 



I carried sealed tubes containing these and other solutions with 

 me, and during several weeks kept the first solution between 

 the temperatures of 26 and 16 , and shook it violently at inter- 

 vals without causing crystals to appear. But as soon as the 

 temperature fell to 16 the solution began to crystallise. 



We conclude, therefore, that in an open vessel such a 

 solution crystallises at 26 only because the liquid is inoculated 

 by crystal germs from the air, and that it crystallises again at 16 

 because fresh crystals appear spontaneously or are born in the 



