THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 109 



Columba hodgsoni presents two interesting points for comparison with Turtur. 

 First, a patch of feathers on the side of the neck have black centers and gray 

 edges; these feathers are acuminate. Second, the feathers of the upper breast 

 and fore-neck are of deep chestnut color, and are broadly edged with pale or 

 gray vinous. 



Columba leucomela 22 is black on all of the upper surface of the body, the wings, 

 and the tail; but the interscapulars, scapulars, and feathers of the whole back are 

 edged with purple in some cases, and with greenish purple in others. This seems 

 to be the turtle-dove pattern only a little disguised. 



Columba laurivora has feathers of the upper breast and fore-neck with dark 

 centers and with lighter edges, these edges being vinous purple of a metallic cast. 



Columba trocaz has the neck-feathers — in the region of the turtle-dove spot — 

 scale-like, with dark-pointed centers and with gray tips or edges. The neck is also 

 iridescent. The bird reminds of the wood-pigeon as well as of the turtle-dove. 



Three species of Columba still preserve the wedge-shaped area, or "guinea- 

 mark," which — as a median extension of the light apical edge — divided the original 

 dark center into two lateral chequers. These are C. guinea," C. maculosa, 2 * and 

 C. albipinnis. In the latter species these marks are apparently confined to the 

 scapulars. 



Finally, there is a group of 5 species which present stages of wing-coloration 

 rather closely approximating conditions found in C. livia. Some have made less, 

 and some more, advance in the transformation than has the rock-pigeon. The 

 species are C. leucoiiota (Tibet), C. livia (Europe, India, etc.), C. rupestris (Man- 

 churia, China), C. oznas (Europe and Turkestan), and C. eversmanni (western 

 Siberia to northern India). 



Columba leuconota stands below the rock-pigeon in having 3 to 4 wing-bars. 

 The two bars of C. livia are represented here as the strongest of the wing, but, in 

 addition, a third and traces of a fourth bar are found just anterior to these, as is 

 shown in plate 50. In the young I find similar bars 25 (pi. 50). 



The broad white tail-band of this species 26 is also present in Columba palumbus, 

 C. oenas, C. guinea, C. eversmanni, C. rupestris, and common pigeons. 2 ' 



Columba rupestris is found in northern China and in Manchuria, and has been 

 described as "intermediate" between C. leuconota and C. livia. Pallas considered 

 it erroneously as a variety of the stock-dove (C. oenas). I have not seen this species, 

 but have had a figure prepared from that of Bonaparte and Pallas (text-fig. 23). 

 It is said to be "much larger" than the rock-pigeon. The posterior wing-bar is 

 wanting in the lower half. The upper half is present, and on the inner (dorsal) side 



22 Called Leucomelana norfolciensis by Bonaparte. 



23 The relation of the guinea-mark (of ( 'olumba guinea and C. maculosa) to the transformation of the dark centers 

 into two lateral chequers has been stated on page 60. Adult, young (and mutants), of C. guinea are shown in color 

 in plates 76 and 77; the neck-mark and its behavior in hybridization are figured and described in Chapter XVII, 

 Vol. II, text-fig. 7. (I find a remark on the relationship of C. guinea and C. palumbus as shown by their tail-bars. Plates 

 were prepared to illustrate the point, but I find neither legends nor further comment on the matter. — Ed.) 



24 Shown in color, plate 18; single feather enlarged, showing guinea-mark, plate 79, figure C. 



26 The complexion and down of the young are as in common pigeons. One of these young was qui'e like an albino — 

 very spare down, and this whitish, with only the faintest trace of yellowish. The beak of this young— unlike the others 

 I have examined — had no black, nor even a brown, bar. (Such light- complexioned and barless-beaked young are 

 also sometimes found among common pigeons. — Ed.) 



26 For head, neck, and eye in color, see plate 80, figure F. 



This tail-band is apparently not found in Columba livia. — Ed. 



