SELECTION 159 



two forms from one, without the help of isolation. 



As this is a very important point, it will be better 

 to illustrate it further. Let us take the simple case 

 of the first organisms, floating on the surface of the 

 sea, and belonging to one species only. We must, 

 however, suppose what was not really the case 

 that they increased in numbers by conjugation and 

 not by self -division ; for, unless we make this sup- 

 position, we cannot reproduce the action of natural 

 selection without the aid of isolation. There are 

 reasons for thinking that the primeval ocean con- 

 tained mineral hydro-carbons, which would slowly 

 oxidise into carbo-hydrates, and thus furnish food 

 for these first organisms. 46 But this supply of food 

 was limited ; and in time it would be necessary for 

 .some individuals to develop chlorophyll, so that they 

 might obtain their supply of carbon from carbonic 

 acid when the carbo-hydrates, were consumed. 

 Other individuals, however, would lead a more 

 animal life and live upon their neighbours. But if 

 free inter-crossing was maintained, these two types 

 could not completely differentiate, for they would 

 always be getting mixed up again, and the whole 

 group of organisms would progress in that single 

 direction which was, on the whole, the most advan- 

 tageous for them. Evidently they could not be- 

 come entirely animal, for they would then lose the 

 power of obtaining food from the mineral substances 

 around them ; and they would not become entirely 

 vegetable, because the power of digesting organic 

 substances, either living or dead, would certainly 



See Chapter II., p. 27. 



