476 



The Journal of Heredity 



associated characters may not im- 

 probably have different relations of 

 dominance or recessiveness from that of 

 their more obvious correlates. The 

 question which we have raised can, at 

 least theoretically, be decided by obser- 

 vation and . experiment. However it is 

 decided will make little practical dif- 

 ference with most of the problems that 

 confront the investigator in genetics. 

 But there are certain problems of 

 genetics, I suspect, in regard to which 

 it will be found to have an important 

 bearing, although its chief importance 

 is in the way it influences our views on 

 certain fundamental problems of onto- 

 geny and evolution. I can here indicate 

 but a few cases in point : 



BEARING ON EVOLUTION 



Since I have come to see more clearly 

 the implications of the question I have 

 discussed, I have been surprised to find 

 how many of the difficulties urged 

 against the theory of natural selection 

 disappear when we consider variations 

 as organismal instead of limited pri- 

 marily to particular parts. Most discus- 

 sions, I find, consider evolutionary 

 problems from the standpoint of the 

 doctrine of unit characters. How com- 

 mon it is to find speculations as to how 

 this or the other character could have 

 been developed through natural selec- 

 tion, as if each part were somehow 

 separately improved by a series of 

 fortunate survivals. If each character 

 is considered as the summation of a 

 series of variations which primarily 

 concern that character alone, and if the 

 nature of the variations that are 

 integrated is determined by natural 

 selection, we should expect most attri- 

 butes of an organism to be of a useful 

 kind. If, on the other hand, variations 



of any one part involve variations 1 , 

 throughout the organism, then the 

 preservation of favorable variations in 

 any one organ would of necessity entail 

 changes in other organs which for the 

 most part would probably have no 

 relation to utility. On this view a 

 considerable ingredient of non-adoptive 

 characters would naturally be expected, 

 and it is probable that, through correla- 

 tion, parts might be evolved to a con- 

 siderable degree of complexity without 

 having any important use in the life 

 of the organism, provided they did not 

 become positively dangerous to their 

 possessors. Much of the evidence ad- 

 dt^qed for orthogenesis is what we should 

 e:< wet to find if evolution occurred 

 through the selection of organismal 

 variations. Much of the difficulty about 

 the beginnings of structures and their 

 development up to the point where they 

 acquire selective value would, I believe, 

 also be removed. The wonder is not so 

 much that selection should produce a 

 large amount of what Haeckel would 

 call dysteleological structures, but that 

 it is able to produce (if we grant that it 

 does produce) so much that is so nicely 

 coadapted, and especially that it is 

 able to carry on the simultaneous 

 elaboration and perfection of numerous 

 separate systems of organs. 



In these days of attack upon evolu- 

 tionary problems through direct obser- 

 vation and experiment, I hope I may be 

 pardoned for presenting anything so 

 atavistic as an academic discussion of 

 the method of evolution. But even 

 with our present accumulation of facts 

 bearing on this much discussed problem 

 there is still something to be gained by 

 reflection, and if our reflection suggests 

 new things to look for it will assuredly 

 not be in vain. 



