180 ORTHOGENETIC EVOLUTION IN PIGEONS. 



no transfer of qualities has taken place, and it is plain that transference or trans- 

 mission is absolutely impossible in the nature of the case. 



It is well to remember that we have in every ontogeny a flowing sequence of 

 specific forms, for every stage is as specific as the end stage. In this progressive 

 change of form we see an interesting difference between the development of the 

 organism and the development of the crystal; nevertheless the form is as certainly 

 a "physical determination" in the one case as in the other. The crystal has its 

 specific form, and sometimes several specific forms. In every case it owes the form 

 to the nature of its material elements and the conditions under which it arises. We 

 would not think of ascribing its form and symmetry to hereditary transmission; 

 neither would we think of intercalating any directing or formative agent, distinct 

 from the material elements composing it. 



Fundamentally considered, the organism and the crystal are equally self- 

 determining at every step, equally the products of intrinsic physical properties and 

 conditions. 



The crystal is said to grow by "accretion," the organism by "intussusception." 

 But this is merely a superficial difference that does not affect the general standpoint. 

 From a physical standpoint the essential thing is not where the elemental particles 

 attach themselves, whether interstitially or superficially, but that they attach 

 themselves in a self-regulating determinate way, so that the typical form at every 

 step is autogenic rather than allogenic (Jensen). 



Now, while developmental recapitulation is so wonderfully exact that we never 

 cease to be amazed at the accuracy of its reproductions, we do not forget that germ- 

 cells are subject to slow variation. Absolute perfection of reproduction could, of 

 course, not be expected, even if outside influences should remain uniform. It is 

 this germ-variation — which we may regard as the error of recapitulation — that 

 becomes so all-important for phyletic departures, for it is the only variation that 

 can be hereditarily recapitulated and thus perpetuated. When the germ-cell begins 

 with a slight initial deviation, the whole series of developmental stages is affected, 

 as we must infer, even though we may be unable to discover the digression until 

 near or at the end of development. Succeeding generations begin recapitulation 

 at the same level, or at the new point in the same level. Through germ-variation, 

 then, recapitulation makes all its permanent advances. 



But how comes it to pass that these advances are, on the whole, adaptive, and 

 progressively so? Recapitulation can only conserve what is given. If it moves 

 onward in a progressive way, there must be some way of limiting germinal varia- 

 tions to lines of cumulative improvement. Here we find ourselves confronted with 

 the difficulty which has long led investigation and theory, and the solution is yet 

 a long way ahead. 



The creation hypothesis may be said to have passed out of science with Louis 

 Agassiz, and to have received its death-blow at the hands of Darwin. The old 

 teleology has lost all standing in science; and vitalism, entelechy doctrines, and the 

 like are no less incompatible with science. Lamarckianism has been effectually 

 silenced by Weismann. The mnemogenesis dreams of Hyatt, Cope, Hering, and 

 Haeckel have not saved the doctrine of the transmission of acquired characters. 

 The " centro-epigenesis " theory of Rignano, suggestive and instructive as is the 

 attempt to approach the explanation of vital phenomena, fails to revive the lost 



