Appendix B. 163 



order to secure survival of the type; but I feel that little 

 reliance can be placed on calculations based on the numerical 

 co-efficient of fertility (i. e. the ratio of the number of offspring 

 to the number of parents) in determining the relative chance of 

 type-survival. 



Take, for instance, the oak tree. It produces thousands of 

 acorns, almost the whole of which die without producing any 

 progeny. Have we any reason to believe that if the number 

 of acorns borne by oak trees were diminished, even so much 

 as to one-tenth, the race of oaks would perish ? It may 

 of course be said that, if all other things are equal, the pro- 

 babilities of survival must be increased by increased fertility 

 of this kind ; but I feel convinced that when numerical fertility 

 has attained to a high point in circumstances in which 

 actual increase of the race cannot take place to any substan- 

 tial extent, the numerical value of this fertility sinks down 

 into a factor of the second or third order of importance that 

 is to say, into the position of a factor whose effects are only to 

 be considered when we have duly allowed for the full effects 

 of all the main factors. Until we have done that, we gain 

 little or nothing in the way of accuracy of conclusion by taking 

 into consideration the minor factors. It may be very well to 

 neglect the effect of the attraction of Jupiter in our early re- 

 searches on the motion or the Moon ; and our doing so will 

 not prevent the results being approximate and having consider- 

 able value, because we are retaining the two main factors that 

 establish the motion, viz. the effects of the Earth and the Sun. 

 But if we exclude the effect of one of these main factors, our 

 results would be worthless ; and it would not be rendered sub- 

 stantially less so by the fact that we had taken Jupiter into 

 account in arriving at them. 



You must not imagine, however, that I think it wholly profit- 

 less to see whether there would be any substantial effect on 

 numerical fertility were selective fertility to manifest itself. But 

 if we want to derive any assistance from calculation, it must 

 be by applying it with a good deal more precision and definite- 

 ness than anything that Wallace shows. And, in the first 

 place, it is useless to confuse the vegetable and animal kingdoms. 

 In the former you have union unaffected by choice ; in the latter, 



M 2 



