THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 103 



feathers with roundish dark spots in a white ground, that these spots are remnants 

 of the original dark center. 



The most anterior feather (No. 1) shows a mere trace of a spot. But the whole 

 center is not lost — it is only the distal third that suffers reduction. If we begin 

 with 1 and pass backward through the whole series to 15, we could say that the 

 gray of the basal part of the feather is gradually increased, and that the steps in 

 evolution led from a dark feather to one with a dark base and a white end; then to 

 the establishment of a dark spot in the white end-segment, and by increase of dark 

 pigment to a coalescence of the basal pigment with the spot, and finally to the 

 turtle pattern in 15. We should then incline to think that the next step would be 

 to complete the series by extending the dark to the very edge of the feather, and 

 thus reach the highest stage of a uniform color. This mode of reasoning has fre- 

 quently misled investigators. 52 



According to my view, the original pattern was the turtle-dove pattern. On 

 the upper parts the light edge of the feathers has disappeared; on the lower parts 

 the original pattern is best preserved posteriorly, and more and more modified as 

 we go forward, so that 1 would represent a higher stage than any of the stages that 

 follow. That this is the correct view can be shown by a general comparative survey, 

 by comparing the juvenal with the adult patterns, and by studying the history of 

 transverse bars in such forms as the flicker (Coloptes auralus, pi. 59). 



Salvadori (page 008) says of the young of this species: " It is similar to the adult, 

 only browner on the wing, and without the dark centers to the feathers of the sides." 

 By "dark centers" Salvadori means such spots as are seen in figures 3 to 8. I find 

 the feathers along the flanks spotted, but the spots are weak and not at all con- 

 spicuous. The middle region of the abdomen and lower surface is whitish without 

 spots, showing a faint trace of buff. The feathers of the side, just behind the legs, 

 have a stronger tinge of buff. 



The color of the young is closely like that of the old bird, but has a plainly deeper 

 tinge of brownish. The feathers of the forehead are a whitish buff, which becomes 

 lighter from the beak over the eye. The under tail-coverts are nearly buff color, 

 and are without spots. 



I find the light apical edge. It is, however, very weak and thin — scarcely noticer 

 able even at close view. It is present in some three or four rows of wing-coverts 

 on the feathers which lie outward (downward) from the middle of each row; above 

 this level (on the inner feathers) it is not visible in these rows, and on the lesser 

 coverts I miss it altogether on all the feathers. 



The secondaries and primaries are pale light buff along the outer edge and 

 around the tips; this pale edge grows less in passing upon the inner webs. Three 

 or four outer tertials have a trace of the apical mark from 0.25 mm. to 0.5 mm. in 

 width. In the long coverts, where these marks are somewhat plainer than in the 

 median coverts, the greatest width of the mark is about 0.5 mm."'" 



"E.g., Hacker, Keeler, and others. (See Chapter VII. — Ed.) 



63 The general color-pattern of the Caleonadins — the last subfamily of the Peristerida? — has not been described 

 by the author. He has noted however, that "the while tail of Catenas nicobarica is a character found only in the adult; 

 in the juvenal stage — called Cat. gmddia: by Gray — the tail is greenish-black (see A. G. Butler: Knowledge Revealing 

 Ignorance, Avic. Mag., Nov. 1910, page 46)." Some will be interested to know that the sentence just quoted seems 

 to be the last statement written by Professor Whitman into his manuscripts. The modified neck- feathers of this 

 species (Cat. nicobarica) are figured in plate 33, Vol. II, and a single feather is herewith reproduced in plate 43, fig. 5. 



Note. — The manuscripts or records used in the preparation of this chapter were found in folders designatedas: 

 follows: A 4, A 0, A 16. Em 1, Em 4, H 2, K 12, Misc. 6, 04; 005, 7, 9, 11, 16, 17, 18, 20; R 1, 12, 13, 16; Sh 24 3 /l 

 W 4, 5, 10, 11, 13; WW 6, XW 2; Z 3, 4, 10, 11.— Ed. 

 8 



