92 NATURAL SCIENCE. Feb.. 



unproved but disproved. On page 73 he tells us, of the "Lamarckian 

 factors," that " No question of value, as selective or otherwise, can 

 obtain in their case." If we grant this, as I think we must ; if we 

 admit that these " factors " involve *' no question of value," but that 

 it is, to say the least, no better than an even chance whether their 

 influence be good, bad, or indifferent, how do modifications produced 

 by them differ from "fortuitous variations " ? 



Those who attribute the opinion that the so-called Lamarckian 

 principles have not yet been proved to be factors in organic evolution, 

 to the influence of Weismann, will do well to remember that we owe 

 to Charles Lyell the demonstration that, until the influence of the 

 conditions of life has been proved to be determinate, until their 

 competency to cause a tendency to progressive development, 

 antecedent to selection, has been made clear, they are of no value 

 whatever as a contribution to the solution of the problems of 

 adaptation. 



It is quite possible that, in the long history of living things, 

 adaptive structures have occasionally been produced by the fortuitous 

 coincidence of fortuitous variations, but the chances against this are 

 so overwhelming that we are justified in demanding demonstrative 

 evidence before we accept this explanation of any adaptation. The 

 production of words and sentences by turning a crank is not impossible, 

 but many generations of readers have approved Swift's statement 

 that this method of advancing knowledge failed to produce a single 

 learned treatise. 



Since all the past history of life is beyond our reach and cannot 

 be made the subject of experiments, and since the scientific study of 

 domesticated animals and cultivated plants is very modern, most of 

 the evidence for natural selection is, and must be, indirect or deductive. 

 Romanes holds, p. 57, that, since there seems to him to be the same 

 sort of evidence of the influence of the " Lamarckian factors," these 

 stand upon as good a logical footing, in the explanation of adaptive 

 structures, as selection ; but the cases are not parallel, and the sort 

 of evidence which is adequate in the one case is totally inadequate in 

 the other case. 



If natural selection acts at all it must result in adaptation ; while 

 the advocates of the " Lamarckian factors " have yet to prove that 

 these '^ factors " can account for any adaptive structure whatever, 

 incipient or otherwise, except so far as this is the result of pre-existing 

 adaptive machinery. 



Satisfactory evidence that an event is the sequence of antecedents 

 which are adequate may be totally unsatisfactory as evidence that it 

 is the sequence of antecedents which seem inadequate. Indirect or 

 deductive evidence may convince us that an adaptation is the result 

 of selection, and may yet be totally unsatisfactory as proof that it is 

 the result of the inheritance of the effect of nurture. Until this factor 

 has been proved to be determinate and adaptive, the proof we must 



