XXI.] POLLEN FIRST BECOMES DEFICIENT. 351 



plants ; but he will recall that I am speaking of 

 ameliorated or domesticated plants, and not of wild 

 ones. In wild plants, the sex -relation is very largely 

 a specialization in each individual case, but in 

 domestic plants this specialization tends to be over- 

 come by the effects of redundant growth -force. 



If it is true that the female sex -elements are the 

 result of constructive or vegetative changes, it would 

 seem to follow that such elements would be most 

 likely to be retained in the great vigorousness of 

 cultivated plants, and that the pollen would first 

 show signs of failing. This is well illustrated in 

 many cultivated species, for deficiency of pollen is 

 by no means uncommon, while good pistils are 

 almost always present. The only important exceptions 

 to this statenient are the double and sterile flowers 

 like the roses, carnations, and snowballs ; but these 

 plants have been bred directly for their doubleness or 

 sterility, and do not, therefore, influence the present 

 inquiry. The berry .-grower knows that all straw- 

 berries have pistils or seed -bearing members, while 

 an increasing number have no pollen. Potatoes now 

 fail to set bolls because the anthers are deficient in 

 pollen, and horseradish does not set seed, probably 

 for the same reason. One who undertakes to per- 

 form experiments in the crossing of cultivated plants 

 soon finds that it is more difficult, as a rule, to 

 obtain good pollen than good pistils. 



An excellent proof that increased amelioration of 

 fruits imposes a severe tax upon the energies of the 

 plant, is afforded by the habitual failure of very 

 many or even the greater part of the flowers upon a 

 fruit tree which blossonjs full. Apple flowers are borne 



