182 ORTHOGENETIC EVOLUTION IN PIGEONS. 



(2) Selection follows chance advantages, and is impotent to guide incipient stages 

 or to explain the presence of non-useful or neutral characters. It enthrones multi- 

 farious variation and refers continuity in progress to external fortuitous control 

 rather than to internal self-regulating processes. Germinal selection (Weismann) 

 recognizes the necessity of "internal regulation," but finds only chance-loaded dice, 

 and selection still supreme. 



(3) Orthogenesis, as here conceived, enables us to predict, to some extent, stages 

 yet to come in the evolution of color-patterns; to trace histories of past sequences; 

 to put order and meaning into fragmentary taxonomic descriptions; and to antici- 

 pate the discovery of elements that have been overlooked. It enables us to under- 

 stand parallel evolution in allied species, even when living under quite unlike condi- 

 tions. It saves us from the stultification of holding selection sufficient to account 

 for those long and definite lines of evolution revealed by paleontology. It helps 

 us to see that biogenetic recapitulation rests in a physical basis and must therefore 

 be taken as an indispensable guide to phyletic sequences, especially in the stages 

 of immaturity, in which allied species frequently differ, not so much in divergence 

 of development as in the extent to which development is carried in one and the same 

 line. That is, two species may differ from each other in the same way that the two 

 sexes in the same species differ, when the male passes on beyond the stage in which 

 the female halts. 



Recapitulation is no myth. Everyone knows that ontogeny repeats its sequence 

 of stages with an accuracy that approaches physical exactness; and that such repeti- 

 tion is attributable to germs of like constitution, developing under like conditions. 

 It is impossible to conceive of such sequences resulting from kaleidoscopic mutations, 

 or from chance-deliveries of any description. The best proof that the organic world 

 is the product of evolution, the first term or terms of which were the units we call 

 cells, is the fact that ontogeny in every species keeps on ever repeating this stage, 

 not as a chance conglomerate, but as a primordium of very definite constitution, 

 capable of developing with clock-like regularity to a specific goal. Consider, too, 

 the remarkable uniformity that prevails in the stages leading up to fully formed 

 germ-cells; the processes of fertilization and maturation; the essential likeness in 

 cleavage-processes, however much obscured by the presence of inert food-yolk; the 

 irreversible sequence of the ascending form-phases, etc. 



In view of all this, what are we to expect in the later stages of closely allied 

 species descended from a single parent form in relatively recent times? Whether 

 we have one-sided branching, equal divarication, or unequal prolongation of 

 the parent line of evolution, the presumption is all in favor of the recapitulation 

 of the ancestral end-phases in more or less typical form in the first or juvenal 

 plumage. 



The immature geopelias offer some convincing illustrations of recapitulation. 

 As an example, we may consider the apical marks of the feathers of Geovclia 

 tranquilla. 



THE APICAL MARK AS AN EXAMPLE OF RECAPITULATION. 



This apical mark, which is seen on the feather tips only for the few weeks during 

 which the juvenal feathers are retained, and then completely vanishes at the first 

 molt, is a character common to all pigeons, if we except a few species that have 



