230 ON VARIABILITY AND ADAPTATION 



ions became connected ? You admit at the start that this 

 original ring and side-chains are immensely complicated, and 

 that our diagram represents an absurdly simple case ; how can 

 you have the face to make any such proposition ? 



All I can say, in reply, is that I have precedent upon 

 my side. What happens in crystallization ? You have a 

 watery solution, say, of common salt. You now know that 

 though you place solid sodium chloride, a definite chemical 

 compound, into that water, in the very act of solution, the 

 ions of chlorine dissociate themselves from the ions of sodium. 

 Molecules of NaCl break up, they are no longer there as such. 

 You now cause this solution to evaporate and become concen- 

 trated ; what happens ? As the water volatilizes do the ions of 

 the gas chlorine escape into the air, leaving the heavier sodium 

 ions behind ? Not a bit of it. Inevitably, when the concentration 

 reaches a certain point, some ions of the sodium, aided by some 

 sharp point or inequality in the surface of the vessel, once more 

 join themselves each to an ion of chlorine, and the process of 

 crystallization begins. The very existence of one crystal clearly 

 seems to cause other junctions to occur in its neighbourhood, 

 and j ust that one particular series of junctions necessary to form 

 sodium chloride, so that, although there may be present in the 

 solution various other substances, various other ions, yet they 

 are not attracted, they remain still in solution to a very great 

 extent ; we gain pure, or almost pure, crystals of the one substance. 

 Here we have a process of the very same order as that which I 

 contemplated above, a process of definite, inevitable, or selective 

 attachment of ions in a certain order, given certain definite 

 conditions. Certain particular ions unite and build up the 

 molecules and crystals of one particular inevitable form, of one 

 particular composition. 



Here again the example is very simple ; I might, it is true, 

 have mentioned solutions of salts of a more complicated nature, 

 but again I hold that the simplest case is good enough, nay, is 

 best, upon which to base my contention, which is that we have 

 to recognize the existence of this affinity or attraction, or what- 

 ever you please to call it, whereby certain freely floating ions, 

 and combinations of ions under certain definite conditions, tend 

 to attach themselves one to the other in a definite order, to 

 form bodies of a definite composition. And in this connexion, 



