THE TURTLE-DOVE PATTERN IN THE PHYLOGENY OF PIGEONS. 97 



According to Dr. Butler (A vie. Mag., n. s., Vol. II, p. 101), the young tambourine- 

 dove (Tympanistria tympanistria) is similarly marked: "All the feathers of the wing 

 and tail arc of a bright coffee-brown color with broad subterminal irregular trans- 

 verse black bands." The figure given in a later volume of the same magazine (Vol. 

 IV, page 308) makes it clear that the young tambourine rises to a stage of irregular 

 cross-bars analogous to what is seen in the young inca-doves (Scardafella iurm and 

 geopclias, in which we find the feathers edged apically with a pale straw-color ( very 

 narrow in inca but conspicuous in geopelias), followed within by a blackish < 

 centic bar, and then one or more quite broken pale dull "bars" (too irregular to be 

 described as bars mere suggestions of bars). In the mature state the pale apical 

 bar is lost and the blackish crescentic bar becomes terminal. 



This form of barring, so far as the black crescent is concerned, is something later 

 in evolution as I have elsewhere (in the present chapter) made clear than the 

 lateral chequers of the mourning-dove, passenger-pigeon, etc. In young geopclias 

 (e.g., pis. 38 and 41 and text-figs. 18 to 21) we have transient lateral chequers 

 in the tertials and longer wing-coverts, and in such continuity with the black 

 crescents that the latter must be regarded as derived from the former. Even in 

 domestic pigeons we frequently see chequers reduced to black crescents. 



In the young inca-doves (pi. 34, figs. C and D) these same transient chequers 

 are recapitulated on the tertials and the long coverts. Although not so black as in 

 the geopelias, they are yet plain and unmistakable homologues. Only two or three 

 of the inner long coverts have this vanishing chequer as a long lateral streak on 

 both the inner and outer edge of the feather. On the remaining feathers of the row the 

 mark appears only on the outer web, and becomes weaker and narrower as we de- 

 scend the row, until on the outer two feathers it is wholly lost. The recapitulation 

 of the marks in the inca and its South American allies, and again in all the geope- 

 lias of Australia even in the diamond-dove, standing at the extreme upper limit 

 of evolution thus far reached in this interesting genus gives us a very important 

 link in the sequence of phyletic stages. 



I venture to predict that the young tambourine, the young cape-dove, and some 

 of their nearer allies will be found to have more or less plain traces of the transient 

 marginal streaks seen in the tertials and long coverts of the inca-dove, and perhaps 

 also dull spots on the outer webs anticipating the spots of the adult. 



Without going into the evidence here, I may say that I have fully satisfied myself 

 that the lateral spots or chequers are derived from the turtle-dove spots, such as are 

 still seen in Turtur orientalis, and, in a somewhat reduced form, in the European 

 turtle-dove (T. turtur). The mode of derivation was by splitting the original 

 central spot into halves. The splitting began at the apex of the feather, a short, 

 wedge-shaped area of lighter color (i.e., reduced in pigment) appearing at this 

 point and extending more and more inward along the shaft, until the divided halves 

 became two separate spots or chequers, more or less pointed at the distal end. 

 The feather thus became double-spotted. Typical " wedge-shaped areas" are not 

 rare in domesl ir pigeons with the chequered pattern, and they are very characteristic 

 marks in the wing of the guinea-pigeon (Colnmba guinea), where they are described 

 as "triangular white pots." They are seen again as a specific character in the 

 spotted pigeon (('. maculosa) of South America. They occur also in the scapulars 

 of C. albipennis of Peru and Bolivia. 



