THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 73 



Running back to Turtur orientalis, we come to entire Deck-heathen :uid fewer rows- 

 six down to only three or four. An extension of differentiation towards the mid- 

 back of the neck has occurred in the derived forms, the two spots here coalescing 

 plainly for the upper three or four rows." 



It is clear that there are two modes of reducing the original black or brown 

 centers: First, by the guinea-mark (the white or light-colored \\ede-shaped exten- 

 sion of the light apical edge) splitting the center into paired chequers; second, by 

 reducing the center through an extension of the whole of the pale feather bonier. 

 It is this second method or extension of method 20 that has been effective in the 

 Turturinae. It is in this way that the somewhat narrowed and pointed centers of 

 the European turtle, the narrow mesial stripes of the tiger and Surate turtles, and 

 the dark shafts of the red ring were formed. Evidence for this latter conclusion, 

 from an hitherto unmentioned source, is easily obtained by hybridization experi- 

 ments. Crossing the Japanese turtle with, for example, the blond ring-dove, which 

 has a whole color no spots except in the collar we obtain in the hybrids feathers 

 whose indications of dark centers are confined wholly to their shafts, or to their 

 shafts and the immediately adjoining parts. 21 



The pale tip or brownish edge of the feather varies in orientalis and furfur in a 

 reverse direction to that of the dark center, the tip being relatively wider and more 

 pronounced in turtur. Although the Juvenal pattern in both of these specie- i- 

 identical in the main with the adult pattern, we may note some difference 



(1) The fulvous-edged feathers in the young include not only the wings, scapu- 

 lars, back, and tail-coverts (as in the adult), but also the feathers of the neck, 

 breast, and even those of the head, which in the adult are not so edged, but are 

 uniform gray-brown. 



(2) The fulvous edges are more nearly straight in the Juvenal plumage, owing 

 to the form of the feathers. 



(3) The fulvous edges are, in general, paler in the young than in the adult. 

 The tertials and posterior scapulars are practically the same as in the adult. 



This light apical edge seems to be universal in birds. 



Other features of the color-pattern of some of the collared doves (Streptopelia) 

 than the neck-mark, already referred to, indicate the direct derivation of these 

 species from a form like Turtur orientalis. 



The 13 species 22 have all lost their dark spots everywhere except in the black 

 collar, and the only remaining visible trace of the spots elsewhere is the narrow light 

 edge shown at the tip of each feather in the Juvenal plumage (pis. 16 and 17, Vol. 

 II). The reduction process has not only washed out the spots, but has also advanced 

 from the darker brown-red color to gray, isabelline, and blond, and finally to pure- 

 white (Streptopelia alba). Albinism is the finishing stage of this reduction process. 



The little red ring-dove (St. humilis) stands nearest to the turtle-doves both in 

 color (pi. 21, Vol. II) and behavior. In this species the red is but an extension 

 and intensification of the light edge of the turtle-dove feather. In its adult feathers 

 the light or pale edge is wholly wanting; but in the first feathers we have a neatly 



" The neck-marks of the several species here mentioned were separately drawn -much enlarged and are shown 

 and described in Chapter XVII of Vol. II. Kn 



20 Thc whole of the pale edge instead of the ontngrnetically older part of it the apical edge here takes part in 

 the invasion of the black. 



"This median streak curiously appears in several of my young pouters (of this year 1910) which have fallen 

 back from a very light gray with white burs to u much darker gray with red bars bordered with a narrow black. 



22 Later (see Chapter XV, Vol. II) Whitman recognized 17 species in this genus. ED. 



