380 THE SURVIVAL OP THE UNLIKE. [XXIV. 



question, the presumption is that varieties propagated 

 by buds wear out sooner than those propagated by 

 seeds, for the experiments of Darwin and others have 

 shown that the especial office of seed propagation is 

 to increase the virility of the species through cross - 

 fertilization. It must follow, therefore, that in the 

 absence of cross-fertilization virility must be less.* 

 (See the note on page 382.) 



But we do not need to consider this phase of 

 the question, for we are concerned with variation 

 (that is, running out) rather than with ultimate 

 longevity (or wearing out). And it is also probable 

 that any tendency towards weakness through lack 

 of fertilization is fully counterbalanced by the pro- 

 tection which such varieties receive under cultivation. 



The question comes simply to this : If buds are 

 taken from parts which possess stable characteristics, 

 they will give stable products under similar condi- 

 tions. But if the buds are taken from parts which 

 have been developed into abnormal conditions and 

 which tend to vary, they must tend strongly to de- 

 part from the parent, especially when the means by 

 which the high development was produced and is main- 

 tained are removed. Bud -variation may, therefore, be 

 said to be indeterminate. The best example of run- 

 ning out in plants propagated by buds is the potato. 

 It is a matter of general observation that varieties of 

 potatoes disappear. Bealt has made experiments 

 which show that in eight years varieties which gave 

 good crops ran out so as to produce nothing. 

 These varieties were grown in the same garden 



* See also Gray, 1. c. 



tRep. Mich. Bd. Agr. 1876, 111, 



