190 ORTHOGENETIC EVOLUTION IN PIGEONS. 



The three figures we have examined show that the wide passage between the 

 juvenal and the adult pattern is no saltation void, but rather a plenum of develop- 

 ment, with transition-phases leading straight from one to the other. We see a 

 juvenal pattern of quite uniform type for all feathers unfolding at the same age. 

 In the oblique tract there is, for some reason, a belated coming of the feathers; 

 but there is no halt in the tide of development, for we see the color-characters are 

 more and more advanced in proportion as the delay is greater, the successive steps 

 marking a definite age-sequence. We infer, therefore, that the intervals between 

 steps may be reduced indefinitely, and the whole series be thus made to approximate 

 a perfect continuum. In that case, we should have to conclude that the develop- 

 mental processes underlying these steps are moving onward continuously during the 

 intervals between steps. 



But do we really require more evidence of this than the normal course of develop- 

 ment reveals? We have seen a very wide gap between two color-patterns largely 

 bridged by steps running in a close progressive sequence. To one who watches the 

 operation from day to day in the several species of the same genus, and in different 

 individuals of each species, and sees that, when all are considered together, the 

 intervals between steps are practically all covered, the evidence of complete genetic 

 continuity seems all-sufficient. To those who see only what can be diagrammatic- 

 ally represented on paper, the short intervals between steps may still seem to be 

 suggestive of diminutive jumps. If we exclude the opportunity for such an assump- 

 tion, through an experimental test, the evidence for continuity will be as complete 

 as we can at present hope to make it. 



The steps between the juvenal (text-figs. 30 to 33) and mature (pi. 39) patterns 

 are obviously transitional. The intervals are conditioned by discontinuities in the 

 issue of the feathers. If the feathers could follow one another in an unbroken 

 sequence, the intervals would disappear and the steps fall into a flowing line. 



We have two sets of feathers to experiment with. We have seen that increasing 

 delay in the time of bringing forth the first feathers results in more and more 

 advanced stages of the color-pattern. We can not readily force such delays in the 

 first plumage; but we can hasten the appearance of the second feathers by removing 

 the first feathers before they are ready to fall by molt. The second feather that 

 appears in the place of a first feather plucked before its time to fall should present 

 a phase in the development of the color-pattern grading towards completion in 

 proportion as its time approaches that of the mature pattern. That it will do so 

 can be experimentally demonstrated. 



The transition phases form a true linear series, ortho in direction and genetic 

 in composition, since there is a common generative bond that makes each step a 

 necessary link in the chain. With the end-term — the dark tip at its full — the gener- 

 ative power rises to its highest level, and henceforth manifests itself only in a 

 regenerative way at each molt. Exhaustion attends every renewal of activity, and 

 in the intervals between molts there is recuperation, but no measurable advance 

 in specific elevation. If renewal of activity be artificially provoked much in excess 

 of the normal requirement, the power may be weakened, so that it will fail of full 

 quantitative achievement. But, however much weakened, there is never a reversal 

 or retreat, in a qualitative sense, from the specific type. 



The intervals between steps may, then, be reduced indefinitely, and the whole 

 series be thus made to approximate a perfect continuum. The developmental 



