ORIGIN OF SPECIES 141 



pear more than once. This improves their chances of 

 becoming incorporated in the species, and if the 

 mutation produces a character that favors snrvival 

 the chance of its becoming estabHshed is still further 

 increased. But it is also not to be overlooked that 

 since most of these mutational changes are not bene- 

 ficial, their recurrence acts as a drag on the race, 

 because in so far as their genes become disseminated 

 they give rise to defective individuals whenever two 

 such genes are brought together. The early death of 

 defective individuals in the wild state may make their 

 appearance less noticeable than under the more favor- 

 able conditions for survival under domestication. 



It is sometimes implied that a mutational change 

 that is dominant has a better chance than a recessive. 

 This is not the case, however, if the dominant charac- 

 ter is neither advantageous nor injurious, but neu- 

 tral. It, then, has the same chance as a recessive. But 

 if the dominant is beneficial, it has a somewhat better 

 chance than a recessive, because, since it comes to ex- 

 23ression from the beginning in the hybrid type, it 

 improves the chances of that type in comparison 

 with the original type. If the dominant is injurious 

 it will be more quickly eliminated than a recessive 

 character that is injurious. 



These theoretical considerations do no more than 

 suggest certain possibilities concerning the theory 

 of natural selection. Before we can judge as to its 

 actual efficiency we must be able to state how much 



