THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 



79 



no dark centers (pi. 30, fig. B). In Ch. pallescens I find the pale edge, the median 

 streak, and spots most nearly of the rock-pigeon type in the juvenal plumage. 

 These are all absent in the adult pattern (pi. 30, fig. C). 



Chamcepelia talpacoti lacks the dark center in the breast-feathers, and has a 

 paler and more uniform coloration (pi. 29) than Ch. passerina. In the young 29 the 

 upper parts are dusky reddish-brown. The scapulars and coverts have a rufous 

 shaft-line, reminding of the mark of young zenaidas. It is noteworthy that the 

 three inner tertials have black chequers on the outer margin, and that only the 

 innermost one has in addition just a bare trace of a chequer on its inner edge — just 

 where we should expect to find the trace or vestige if any remained. These chequers 

 do not lie cross-wise or obliquely on the feather, but are pointed directly backward 

 and follow along the feather's outer margin, leaving only a very narrow pale edge 

 or boundary-line continuous with the apical mark. On the four inner long coverts 

 the chequers are similar, though shorter, and wider at their anterior ends. These 



Text-figure 15. — Showing the oblique streak in a common pigeon of 7 weeks and 2 days. This region — of the 

 lower, anterior wing — is unshaded; the feathers of this area are gray, without the mixture of brown seen else- 

 where. In this same region the adult color is quickly attained, x 0.7. 



are again located on the outer web, and I fail to see any vestiges on the inner webs. 

 The shape of all these spots or chequers of juvenal feathers is more like the typical 

 rock-pigeon chequers than are the spots of adult feathers. 



In the genus Peristera we find the wing of P. cinerea bearing numerous spots 

 (pi. 31, fig. A), and some of the anterior lesser coverts showing the dark centers 

 (pi. 31, figs, la-9), though these are less strong 30 than those of ( 'hamcepelia passerina. 



In the genus Columbula a wing-pattern of some interest is found. The wing of 

 C. piqui bears anteriorly, in the oblique tract, only an isolated series of conspicuous 

 chequers that might seem to suggest a character quite new in this field (pi. 32, 

 fig. A). Glancing at conditions in some related genera — Chamcepelia, Peristera, Geo- 

 pelia, and even in Zenaidura" — it becomes clear (pis. 32, 33) that these latter reveal 



29 The young described were out of a male from Brazil and a female supposed to have been trapped at Vera Cruz. 

 See note to plate 29, figure A. — Ed. 



30 Among a group of birds received from Brazil is one in juvenal plumage which I should call Peristera cinerea. 

 The three inner tertials have each a pale spot on the outer web; the other tertials have no spot. The first four inner 

 long coverts are spotted, with the fifth unspotted. The first five median coverts are spotted, with l lie sixth unspotted. 

 < >ne lesser covert from about the middle of the oblique streak has a fairly well defined spot. The spots in most of 

 these feathers are remarkably shadow-like, especially on the median coverts and tertials. There is a median buff 

 freckled streak in each of the median and long coverts. There are buff apical marks, and a faint trace of a subapical 

 dark bar. 



31 It seems that no special illustration was prepared of the oblique bar in Zcnaidura. — Ed. 



