8 ORTHOGENETIC EVOLUTION IN PIGEONS. 



When, then, an evolution of millions of years has to be repeated in a few days 

 or weeks, there must be enormous abbreviation in earlier stages, relieved here and 

 there by a few essential landmarks, and only in the later stages relatively fuller 

 exhibition of details. If we are fortunate enough to find a case here and there 

 where lost history can be recalled and actually pictured forth in the living bird, 

 we have, as I have said, a demonstration of a physiological law in development that 

 must have universal validity. The process is not less one of recapitulation or 

 repetition for being condensed or abbreviated. 



These surface-characters, then, that at first sight seem so bewildering in their 

 profusion, so baffling in their sudden seasonal and nuptial transmutations, so 

 puzzling in their age-sequences, turn out to have quite peculiar advantages for 

 the student of evolution. Their perfection as specific marks, delicacy as tests of 

 variability, accessibility in breeding and hybridizing experiments, the convenient 

 serial disposition of their intergradations in rows of feathers, the phyletic sequences 

 they reveal in successive plumages, the ease with which we can force them to bridge 

 gaps in normal development and reproduce lost stages, all these are advantages 

 that speak for themselves. 



In some birds there are several successive patterns graduating up to the adult 

 pattern through as many molts, partial or total. In such cases it is often possible 

 to find practically all the transitional phases that connect the earlier with the later 

 pattern. Here nature turns the molting process into an experiment essentially the 

 same as that I have tried on the diamond-dove and other geopelias. Whenever a 

 bird develops its feathers slowly, so that they appear at different ages in different 

 regions, as in the jungle-fowl, quails, pheasants, etc., it is easy to see that the pas- 

 sage from stage to stage is not by mutations, but by continuous development. 



The testimony from the avian world in respect to the evolution of color-patterns 

 is, as I have said, often fragmentary, and herein is to be found the only apology 

 for an appeal to the theory of mutation. Wherever the evidence is fairly complete 

 and wherever we find it possible to bring the apparent breaks in continuity of 

 development to a crucial test, there we find positive confutation of the mutation 

 hypothesis. 



The evidence for mutation is mainly of a negative character. New species arise, 

 and we discover them when it is too late to learn the history of their genesis. We 

 search through the biological world for discontinuities, and we find them so numerous 

 that ponderous volumes can not approach a full record of them. In our breeding- 

 experiments, with both animals and plants, we not infrequently get departures from 

 the parent type, and these unexpected novelties, seemingly in contradiction to 

 the law of hereditary continuity, are then declared to have come suddenly into 

 existence, to have slipped the bonds of parentage, to have hoisted themselves 

 into existence inconsequentially, per saltum. Shall we assume a break in the chain 

 at every point where our knowledge fails? or is it more rational to abide by the law 

 of genetic continuity so long as we fail to discover evidence of an exception? 



