Hybridization 99 



may grow, but the plants they produce may be wholly 

 barren, sometimes even refusing to produce flowers 

 as well as seeds, as in the instance of some hybrids be- 

 tween the Wild Goose plum and the peach. Sometimes 

 the refusal to cross is due to some difference in the time 

 of blooming or some incompatibility in the structure of 

 the flowers. But it is enough for our purpose to know 

 that there are certain characters in widely dissimilar 

 plants which prevent inter-crossing, and that these 

 characters are just as closely and just as much influenced 

 by change of environment and natural selection as are 

 size, color, reproductiveness, and other characters. 



The limits of crossing tend to preserve the identity of 

 species. Here, then, is the sufficient answer to the prop- 

 osition that inter-crossing must swamp all natural 

 selection, and also the explanation of the varying and 

 often restricted limits within which crossing is possible. 

 That is, the checks to crossing have been developed 

 through the principle of universal variability and natural 

 selection, as has been shown by Darwin and Wallace. 

 Plants vary in their reproductive organs and powers, 

 as they do in other directions; and when such a varia- 

 tion is useful it is perpetuated, and when hurtful it is 

 lost. Suppose that a certain well-marked individual of 

 a species should find an unusually good place in nature, 

 and it should multiply rapidly. Crosses would be made 

 between its own offspring and perhaps between those 

 offspring and itself in succeeding years; and it is fair 

 to suppose that some of the crosses would be particularly 

 well adapted to the conditions in which the parents 

 grew, and these would constantly tend to perpetuate 



