DEVELOPMENT BY NATURAL SELECTION. 115 



tion. But when these laws have to pass, one by one, 

 before the bar of Reason, we see that they have only 

 the limited validity of provisional working hypotheses, 

 although they are indispensable in helping us to bring 

 order and prevision into our perceptions and actions. 

 In other words, they are only ideal fictions. When, 

 therefore, it is argued that the distinctive conception 

 of life is not in accordance with other " laws of Nature " 

 derived solely from past observation of the inorganic 

 world, we are justified in dismissing the argument. 



The contention that it must be possible to predict 

 the phenomena of life from the properties of the chemical 

 substances composing the organism must be set aside 

 at once, since the substances we can separate from the 

 living body are evidently altered in behaviour by the 

 separation. From the chemical standpoint a living 

 organism might, perhaps, be regarded as one huge 

 molecule or chemical system of molecules with easily 

 dissociable parts, each containing innumerable unstable 

 side-linkings, constantly taking up and giving off 

 smaller molecules of different sorts from or to the 

 environment. Except from observation of living or- 

 ganisms, we have no knowledge to guide us with regard 

 to the behaviour of such a molecule or the various 

 atomic linkings contained in it. Nor have we any 

 means of predicting, except by actual observation of 

 such a molecule, the liberations or absorptions of free 

 energy associated with the building-up and breaking- 

 down processes in the molecule. All that we ascertain 

 by the measurements which verify the principle of 

 conservation of energy in the living body is that when 

 the organism, or the molecule, has deviated from, and 



