THE PROBLEM <>K TIIK OKKilN <>K M'l.i , 43 



This simple mode of transformation is ;ilx>ut three-fourths completed in the 

 upper feather of the row, about half completed in the second feather, and only 

 just begun in the third feather. If the figure (te\t-fig. 10) tells the truth, there can 

 be no mistake about the conclusion. The direction and the manner of change an- 

 what we might have anticipated (compare also wing of adult female, text-figure I I I. 



But is not this "bridge" a thing of chance that may appear once in a thousand 

 times, more or less? And may we not find it next time in another place ami telling 

 a different story? Kven the upper spot here is not quite reduced to a typical band, 

 and may it not be a mistake to identify it as a stage of transitional significance? 

 For certainty on this point we require a longer series, with the steps ending in a 

 finished band. We should find the series invariably in the same place and running 

 always in the same direction, in both sexes and at all ages. All these desiderata 

 are amply fulfilled. 



The color-pattern does not reach full development in the first plumage and the 

 male carries the development somewhat farther than the female. The female 

 presents two and the male presents three consecutive stages in this development. 

 Although the stages run nearly parallel in the two sexes, the female not only lias 

 one stage less than the male, but her two stages represent somewhat lower grade.- 

 than the corresponding stages of the male. Hence we have the transitional serie.- 

 in five stages, which ascend in the following order: (1) juvenal female ipl. 1 I ; 

 (2) juvenal male (text-fig. 10); (3) adult female (pi. 15); (4) adolescent male (first- 

 row coverts compared in pi. 17); (5) adult male (pi. 16).' 



Two different sequences are here to be distinguished. One of these falls within 

 a single row of feathers in the individual plumage and is represented in the series 

 of transitional steps leading from the new into the old character. Its closeness 

 and length vary considerably from sex to sex and according to age, but only slightly 

 in different individuals of the same sex and age. The second sequence runs parallel 

 with the development of the bird, from the stage of the first plumage to that of 

 full maturity. It is not a sequence of transitions, as presented in a single row of 

 feathers of a single plumage, but a sequence of the stages of the same transitional 

 series, as presented in successive plumages. The accompanying diagram >. text- 

 fig. 12) will illustrate the relations of the two sequences in the anterior bar in tin- 

 three plumages of the male. 



The single feather with its transitional mark is the unit of the first sequence; 

 the transitional series as a whole is the unit of the second sequence. The first 

 sequence has continuity serially stamped upon it by nature; the second sequence 



6 Before- we go on with the examination of these stages we must take n.it e < if the new elm met rr which invades the 

 Held occupied l>v them. This new character is a brilliant metallic luster a bron/y green in the long coverts, where 

 we meet the transitional phases, and a metallic purple in the secondaries or last row of wine feathers. This r 

 evidently one of the latest acquisitions of the species, is increasing, and spreads from the tip of the feather inward, in 

 a direction precisely opposite to that of pigment reduction. It appears first of all in the field of the Mack s|Mit. but 

 spreads inward beyond the limits of the spot. The outline of the spot disappears in the full bl.-i/.e of i> 

 but it may be readily brought into view by turning the feather until metallic reflections no longer re.ich the e\c Tin- 

 can be done most easily with feathers mounted in serial order on a card-board. Held at the proper angle, the outlines 

 of the spots become clear, and we have before us a full series of transitional phases. The sequeii' 



11 continuous whole in which the stages are merely favored terms in a flowing series of intergrades. Iridescence thus 

 appeals to be a phenomenon tending to elevate the spots and bring them within the sphere of utility. It seems not 

 only to put a check upon the reduction of pigment, but also to actually turn the tide in the opposite direction, for the 

 reduction in this region is not carried so far in the old as in the young male and the femal> 



As the acquisition of metallic brilliancy is accompanied by an exceptional love of display in the male, the chief directing 

 factor in its development may well be natuial selection. 



