54 PHYLOGENY 



and the white streak along the edge of the wing mark a plane in 

 the evolution of this bird very nearly identical with that of the 

 white-winged pigeon. A little higher plane has been reached by our 

 band-tailed pigeon of the Pacific coast, which is also a species of 

 turtle-dove derivation, as shown in the neck-marking and in the 

 voice and behavior. 1 



These illustrations, which could be extended into the hundreds, 

 may be concluded with two cases, representing wide extremes, yet 

 governed by the same law of progressive orthogenetic variation. 



The crested pigeon of Australia (Ocyphaps lophotes) stands at 

 one extreme, at the uppermost limit in the number of bands and in 

 the perfection of finish. There are eleven, or at most twelve, parallel 

 bands crossing the wing and scapulars transversely, each band 

 marking a single row of feathers, with the regularity of zebra stripes. 

 The width of these bands increases from before backward, begin- 

 ning with a width of about \ mm. and reaching 4 to 5 mm. on the tenth 

 band. The eleventh band, located on the long coverts, is especially 

 interesting, as it begins above with narrow elements, like the pre- 

 ceding, but is continued, from the third or fourth feather onward, 

 by elongated checker-like spots. - 



This band, or bar, is the homologue of the anterior bar iii the rock 

 pigeon, and furnishes a standing picture of transitional continuity 

 from one character to another, and at the same time settles beyond 

 dispute the direction variation has pursued. So clear and decisive 

 is the case, that one might safely predict that this entire bar is 

 destined to be reduced to the narrow band-type seen in the fore part 

 of the wing. We have only to turn to a closely allied species, the 

 white-breasted crested pigeon (Lophophaps leucogaster) 2 to find that 

 it has already realized the prediction to the full, having every checker 

 in this row converted into a typical band-element. 



Moreover, the transformation has already begun in the first 

 feather of the next and last row, so that the same prediction could 

 be extended to this bar, which is the homologue of the posterior 

 bar in the rock pigeon. 



Glancing again at Ocyphaps, and looking at the wing as a whole, 

 the course of transformation, its mode, direction, and future term- 

 ination are all very clearly defined. The wing-pattern, as shown 

 especially in the light edges of the Juvenal plumage, takes us clear 

 back to the turtle-dove type. Next came the checkered pattern 

 similar to that of the primitive rock pigeon. Reduction of pigment, 

 proceeding from before backward, fashioned the bilateral checkers 

 from the uni-central spots. The reduction kept on in the same 



1 Minute blotches of black were found in the longer scapulars of a few individuals. 

 These are probably atavistic reminiscences of lost spots. 



2 This bird is comparatively rare, and I have seen but a single pair that recently 

 came to hand through the kindness of Frank M. Chapman. 



