THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OP PIGEONS. 105 



(page 89) notes that "this species passes through many stages before assuming the 

 full dress." The young is so different from the male that some have found it diffi- 

 cult to believe it belongs to the same species. The young, however, does resemble 

 the adult female. The wing is green, but the smaller coverts are edged with a lighter 

 bronze (?), presenting an appearance that closely resembles, in general effect, the 

 wing-pattern in the turtle-dove. The dark (green) centers are more or less pointed, 

 especially on the larger scapulars and inner tertials. The long coverts have green 

 centers and are edged with yellow, and the same is true of the secondaries. The 

 breast-feathers have green centers and light-grayish edges or tips; the upper breast- 

 feathers are bifid and bear whitish tips; the feathers of the lower breast have green 

 centers and wide grayish-yellow edges, just the turtle-dove type of pattern. 



5 



THE TURTLE DOVE PATTERN IN THE GOURIDj<E. 



(See table 1, page 60.) 



The important differences which separate the Gouridre from other families of 

 pigeons have been stated by Goodchild 6 as follows: 



"(1) In the normal pigeons an oil gland is present, but it is absent in Goura. (2) In 

 the normal pigeons the tail feathers are 12 in number, while there are 16 in Goura. (3) In 

 the normal pigeons the pterylosis is columbine, but it is galline in Goura (!). (4) In the 

 normal pigeons caeca are present, but are absent in Goura. (5) In the normal pigeons a 

 gall-bladder is present, no gall-bladder in Goiira. (6) In the normal pigeons incubation 

 lasts 16 days, but extends to 28 days in Goura." 7 



I have seen but two individuals of this group — two Goura coronata, obtained 

 from a dealer in Antwerp. The white color of several of the greater wing-coverts 

 and the variety of the wing-colors in these birds indicate that their color-pattern 

 is a considerably modified one. In the anterior wing-coverts, however, we find 

 (pi. 45) feathers with centers of dark slate, broadly edged with chestnut. The 

 broad edges and pointed centers clearly suggest the conditions found in the European 

 turtle-dove. 8 



THE TURTLE-DOVE PATTERN IN THE COLUMBID/E. 



(See table 1, page 66.) 

 Feral Columba.' 



This very large genus, according to Salvadori, includes 58 species. This author 

 says (page 241) : 



"I feel quite sure that the numerous species of the genus Columba ought to be arranged 

 in several subgenera; but as only some of these can be easily defined, while others defy, 

 at least for the present, precise definition, I have thought it best to leave them all, as a 

 whole, in one genus." 



5 Of the subfamily Carpophagime the author seems to have had in his aviary only a single specimen. An Oceanic 

 fruit- pigeon (Globicera paeijica) was kept in captivity for a period of IS months, but it seems not to have been figured 

 or described. — Ed. 



8 Proc. Zool. Soc, 18SG, page 193. 



7 Huxley (Proc. Zool. Soc., 1868, page 302) noted the aberrant vertebral formula of Goura coronata. This bird 

 is about 30 inches in length. 



8 The Didunculidff' — the fifth family of pigeons — is now represented by a single species, which h as been figured 

 by Bonaparte (Icon. Pig., pi. cxl). Salvadori (page 626) states that the young of this species is entirely brown, but 

 the feathers are edged with "crescentic rufous bands." 



» Neither of the 7 known species of the other 3 genera of the subfamily Columbina: was studied, figured, or 

 described. — Ed. 



