No. 610] GENETICS VERSUS PALEONTOLOGY 627 



anatomy, geology, phylogeny, etc., are practical arts which 

 have to be learned by exi)erience. Phylogenists must con- 

 stantly distinguish between primitive and specialized 

 characters, and if their experience, caution and judgment 

 be adequate they may be as successful as physicians are in 

 diagnosis. Of course physicians make mistakes and so 

 do phylogenists, but in the long run both succeed in sifting 

 the false from the true, even without the aid of direct ex- 

 perimentation. Nature herself often provides control ex- 

 periments, as when she forces animals of widely different 

 stocks into similar life habits, or when she takes a prim- 

 itive type of skull and dentition and molds them into a 

 wide variety of adaptive types, meanwhile preserving the 

 original pattern as a ''control," either in the form of a 

 "living fossil," persisting in a primitive environment, or 

 in the form of a real fossil found in Tertiary strata. 



Professor Morgan makes a serious and important criti- 

 cism of the comparative anatomical and paleontological 

 doctrine that structures have been derived by progressive 

 continuous stages. He is evidently inclined to think that 

 sti-uctures have rather been derived through discontinuous 

 mutational stages. It would be easy, he shows, to arrange 

 a graded series of fruit flies belonging to distinct muta- 

 tions, having at the one extreme perfectly formed wings 

 and at the other extreme no wings at all. But this series 

 by no means represents the historical order of appearance 

 of these mutants, which are not genetically derived one 

 from the other, but have arisen independently. Again 

 (p. 13) 



which is deep brown, to pick out a perfectly graded series [of races] 

 ending with pure white eyes. But such a serial arrangement would 

 give a totally false idea of the way the different types have arisen; 



well be^entirely erroneous, for the fact that such a series exists beare 

 no relation to the order in which its members have appeared. 



''Suppose," he continues, "that evolution 'in the open' 

 had taken place in the same way, by means of discontin- 

 uous variation. What value then would the evidence [for 



