THE EVOLUTION OF LIVING BEINGS. 21 



experiment. So he began to examine the flora of France 

 carefully and was soon able to show that differences 

 were observable within all Linnean species, that whate- 

 ver Linnean species is carefully observed in nature, it can 

 be dismembered into groups of individuals with com- 

 mon characters, differing from other groups of indivi- 

 duals with different characters, but also common to 

 these latter ones. 



The only logical conclusion so far was : the Linnean 

 species is not a unit, but a mixture of individuals of 

 different constitutions. 



But this conclusion had to be proved. 



How to do this? 



The closest examination, Jordan quite seized this 

 point, would be insufficient to test the constitution 

 of an individual, because if such a thing as variability re- 

 ally did exist, there was no reason to suppose that an 

 individual, modified in an unheritable way, could be 

 distinguished from an individual, possessing the same 

 characters innately, and consequently in a transmit- 

 table way. 



Jordan consequently discarded morphological compari- 

 son as a criterium for specific purity and, falling back 

 to Ray, (of whom he may or may not have known) sub- 

 stituted for it : nulla certior. . . . quam distincta propaga- 

 tio ex semine. 



This was a great advance, a milestone on the road 

 to knowledge. Carefully collected seeds, each lot from 

 a single individual, protected against contamination 

 by crossing, brought, upon sowing, the important fact 

 to light thai J within the Linnean species there are indi- 



