46 PHYLOGENY 



or character begins with an "infinitesimal rudiment/' which has no 

 way of emerging from its functionless state except through minute 

 chance variations in various directions. In this assumption the 

 problem is misconceived. The characters we meet with to-day have 

 rarely, if ever, arisen by direct evolution from useless rudiments. 

 When we know enough about a character to undertake to trace its 

 genesis, the "rudiment" imagined to lie so near recedes, and we 

 are led on, not to a "beginning," but to an antecedent; and if we 

 are fortunate enough to be able to advance farther, we come to 

 another antecedent, and so on. The series of antecedents stretches 

 ever as far as we can see. As we repeat this experience with different 

 characters, looking always for the primordial rudiment, our childish 

 faith in such "beginnings" gives way to the conviction that the chase 

 is led by a phantom. 



No one of our sense-organs, for example, can be traced to a rudi- 

 ment, except in the embryological sense. The eye of the vertebrate 

 may appear as a rudiment in the embryo, but no one can doubt that 

 it has had a phylogenetic history, the first term of which if first 

 there be must have been very different from its present embryonic 

 rudiment. To assume that the eye began in some indifferent variation 

 that fluctuated or mutated, chance-wise, into a state of incipient 

 utility, and was then developed in a direct line to its present stage of 

 complex adaptations, either gradually or per saltus, would be hardly 

 more satisfactory than appealing to a miraculous succession of 

 miracles. It is impossible to believe that such a system of har- 

 monious coadaptations could ever arise by mutation; 1 and selec- 

 tion, although playing a very important part in such achievements, 

 is probably never equal to the whole task. Without the assistance of 

 some factor having more continuous directive efficiency, selection 

 would fail to bring out of the chaos of chance variation, or kaleido- 

 scopic mutation, such progressive evolution as the organic world 

 reveals. 



In order to show that such a factor is essential, and that it is 

 actually present, supplying the indispensable initial stages, and 

 holding the master hand in the general direction of evolution, demon- 

 strative evidence is, of course, required. Such evidence lies in the 

 history of specific characters. But how shall we approach such a task, 

 if no near-by rudiment is to be found as a starting-point? Rudiments 

 and premutations are alike illusory in this regard, for their beginning 

 is always and necessarily assumed to lie in the realm of the invisible 

 and unknowable. If we are to keep always on ground that is open to 



1 Darwin frequently emphasized the same objection. In a letter to Asa Gray, 

 referring to the orchids, he remarks: " It is impossible to imagine so many coadap- 

 tations being formed, all by a chance blow." 



Weismann has shown in a masterly manner how inadequate is the mutation 

 theory to account for such phenomena. 



