BASIS OF THE PKOBLEM ^75 



the influence of the environment, the mutations themselves may 

 form a series separated by steps that are so small as to be scarcely 

 measurable or which may be very large. The weight of the beans 

 is an example of the former kind of mutation ; the so-called 

 ' meristic ' variations, when another member is added to a series, 

 the addition, for instance, of a vertebra to the vertebral column, 

 is an example of the latter kind. 



Whenever a mutation occurs, we have to think of it as founded 

 upon some change in the germinal constitution. Such changes 

 are of the nature of modifications of factors — and may be positive 

 (leading to the apparent addition of factors), negative (leading to 

 the apparent loss of factors), or qualitative. But of the nature 

 of these changes we know" little, and of their causes less. What 

 is important, however, is that these changes do occur. Further, 

 they occur in all directions. The direction is, of course, in a sense 

 determined by the starting-point — by what is already given in 

 the germinal constitution — but, given the starting-point, muta- 

 tions apparently occur in all directions. It is a matter of impor- 

 tance to know whether mutations ever tend to occur more in one 

 direction than in another. Nothing definite has been ascertained 

 as to this particular problem, though, as we shall note later, 

 certain facts with regard to the evolution of animals suggest that 

 there have been tendencies to change along certain lines. It may 

 also be asked whether the continued selection of a character in 

 any way affects the direction of the variation of the germinal 

 constitution. To this again there is as yet no definite answer. 

 On the whole it is not probable that selection has any such effect. 



8. So far, therefore, as we have at present gone, we have found 

 that permanent change is of the nature of germinal change. We 

 have also discussed the nature of germinal differences as they 

 exist between different individuals, and we have discussed the - 

 nature and causes of germinal change. With regard to variation 

 there is much which is doubtful and obscure, but bearing in mind 

 the essential features of what is certainly known, we may turn 

 to consider what passes in the organic world. Upon this subject 

 there is much less uncertainty. We have seen how huge is fecun- 

 dity. Of the young in any generation, only about that number 

 survives which equals the number of adults in the generation 

 from which they are derived. For normally in the organic world 

 the total number of adults of any species at corresponding periods 



