1 92 PROBLEMS OF EVOLUTION 



contrary, her method of working is well calculated to swamp 

 each new variety before it is worthy to rank as a species, 

 inter- But this result does not follow since the most closely allied 



between s P e cies among wild plants and even varieties within species are 

 species sterile inter se. Romanes was the first to realise the importance 

 of intersterility. On it he based his theory of Physiological 

 Selection, which I have already discussed. It is impossible to 

 hold with him that a variation in the reproductive cells will give 

 rise to new species if unaccompanied by some other variation 

 that gives the plant an advantage. All I contend for is that 

 favourable variations in flowers have been preserved from the 

 swamping effects of intercrossing by the sterility of the deviating 

 plants when crossed with those of allied varieties and species. 

 That such sterility has played an important part in the evolution 

 of plants can hardly be doubted. A number of French botanists 

 made experiments on this subject, continued through many years, 

 with the object of proving that evolution was a myth. 1 But by 

 the irony of fate the results of their experiments came in oppor- 

 tunely to support Darwinism. These botanists crossed a number 

 of varieties distinguishable only by the most minute differences, 

 and found that the transferred pollen had no fertilising effect. 



The bees themselves are always making experiments of the 

 same kind on varieties and closely related species, and in spite of 

 them even many varieties keep true. We are not, therefore, 

 dependent for our facts on the savants in question. There is 

 other first rate evidence that any man can verify for himself. 



Varieties This being the case, it is curious that varieties of garden 

 flowers are not characterised by the same intersterility, and require 

 careful isolation, if seedlings are to be like the parent plant. 

 The explanation of this, probably, is that among wild flowers, 

 of all the countless varieties that have arisen, only those have 

 survived in which the advantageous peculiarities have been 

 accompanied by intersterility with nearly allied varieties. 



Among wild animals it is probable, as I have said, that inter- 

 sterility between species is not so general as it is among wild 



1 An account of these experiments is given by Romanes in his Dartvin and after 

 Dariuin, vol. iii. pp. 86-89. 



