54 ORTHOGENETIC EVOLUTION IN PIGEONS. 



Australia. On the secondaries and the longer coverts these chequers fall into regu- 

 lar rows, conforming necessarily to the serial arrangement of the feathers bearing 

 them. They are thus in a position and order corresponding precisely to the position 

 and order of the elements of the bars in Columba livia. To convert them into 

 typical bars it would be necessary only to enlarge them transversely until they 

 coalesced by juxtaposition. These transient chequers, distributed to all the coverts 

 of the wing, seem to tell us that the geopelias were descended from chequered 

 ancestors, whose color-marks are now recapitulated, but soon superseded by the 

 later pattern which characterizes the species in the adult stage. The barred con- 

 dition of the adult plumage differs widely from that of C. livia; but the important 

 point is that longitudinal chequers are first and transverse bars second in order 

 of development. 



While all species of Geopelia exhibit the same transitory chequers in the juvenal 

 feathers, in the adult plumage two distinct patterns are presented. In G. cuneata" 1 

 the ground-color is gray and each of the upper wing-coverts is marked with two 

 small lateral, roundish, white spots, which fall into regular transverse rows on 

 the longer coverts. These rows cross the wing in lines that would be followed by 

 transverse bars, and development shows that the spots are differentiated remnants 

 of bars. 



In the other geopelias, narrow transverse bars are formed by the coalescence of 

 the dark crescents at the tips of the feathers. As these bars follow the tips of the 

 feathers, they are really composed of serially connected scallops that simulate bars 

 only in a few of the longer rows of coverts, where the feathers overlap in a way to 

 conceal the upper third of each crescent and leave the lower third exposed as com- 

 ponent parts of nearly straight and continuous bars. 



In the ontogeny of the geopelias we see the different species all taking depart- 

 ure from a common color-pattern, consisting of longitudinal spots or chequers 

 distributed uniformly to all the feathers of the wing. We see this common and 

 earlier pattern retained for only a few weeks, then exchanged for the specific 

 patterns of the adult birds, the later patterns developing in two widely diverging 

 directions, leading to two types of generic rank. 



In the geopelias the chequers are completely lost with the juvenal feathers; 

 in a distinct but allied genus (Ocyphaps), the crested pigeon, an interesting 

 variation is presented. 8 Here typical chequers have been retained in two rows of 

 feathers the long coverts and the secondaries and replaced in all the other upper 

 wing-coverts and scapulars with narrow transverse bars, differing from those of 

 Geopelia in being subterminal and straight and in being directly derived from the 

 original chequers by simple reduction of the latter in length. In the second row of 

 coverts next in front of the long coverts the reduction is less than in the 

 anterior coverts, and consequently the bars are here more evidently shortened 

 chequers. In the young bird the bars shorten gradually from behind forward, from 

 chequers of full length to the narrowest anterior bars. The mode of trans- 

 formation is demonstrated to perfection in the two inner feathers of the row of 

 long coverts. 



7 Reichenbach gave this unique species the generic name Stictopelia, and Bonaparte concurred in this. I think 

 the development of the color-pattern warrants the separation from the other geopelias. 

 This is fully described in the preceding chapter. ED. 



