112 



BEHAVIOR OF PIGEONS. 



ON SPECIES RELATIONSHIPS IN PIGEONS. 



The main stem of the pigeon branch, if I read correctly the testimony of color-patterns, 

 is represented most nearly by the turtle-doves (Turtur orientalis and Turtur turtur) of the 

 Old World. The original turtle-dove pattern, in which all or most of the feathers were 

 similarly differentiated into a "dark center" and a "light edge," seems to have been a 

 very general if not a universal avian pattern. This pattern certainly preceded the chequered 

 type of the rock-pigeon (C. affinis), and it is possible still to find connecting types types 

 in which the turtle pattern coexists with the chequered pattern, the latter coming in to 

 replace or supersede the former. Examples are to be seen in the bronze-winged pigeon 

 (Phaps chalcoptera) of Australia, the Florida ground-dove (Chamospelia passerina), and 

 some others. (EM 1.) 



Ectopistes (passenger, 

 snarls when angry and 

 flirts-wings like Turtle ; 

 Columba. does not). 



From Turtle-dove type 

 -"- | --^ 



Nesopelia ( Galapagos sp.) 

 Zenajdura (mourning dove) 



Zenaida 

 Melopeha (white -wing ) 



Chamaepelia(ground-dove) 



Wood pigeon 



Japanese Turtle 



European Turtle 



Chinese and Asiatic 

 Turtles 



Red Ring 

 Japanese Ring 

 Common Rings 

 (blond and white) 



Turtle type 



(Inca dove) Scardafella 

 Geopelia 



(Cuneata) 



(Humeralis) 



FIGURE 2. 



^" Phaps (bronze-wing) 

 Ocyphaps (crested pigeon) 



In the above, parts A and B were drawn by Wallace Craig, part C by Oscar Riddle. 

 Of the genus Columba, C. palumbus seems nearest to the turtle-dove. C. fas data seems 

 separate from the others. C. rufina is probably close toflavirostris. C. squamosa is probably 

 higher than C. leucocephala. The stock-dove (C. oena) and the rock-dove (C. livia) are 

 more closely related to each other than to the wood-pigeon (C. palumbus) ; the stock-dove 

 probably came, not from the rock-dove, but from the same ancestor. C. gymnopthalma is 

 related to C. squamosa and C. leucocephala. 



One of us (W. C.) often asked the author to state his conclusions as to species 

 relationships; he never replied at great length. The following notes, made at the 

 time, present some of his conclusions and are worthy of record : 



"The Japanese turtle-dove (T. orientalis) is the most nearly ancestral. From the Japanese to 

 the Indian and thence to the European species there is a continuous transition toward smaller size. 

 There is a similar transition in ring-doves from Japan to Europe. 



"The Inca-dove is probably related to the geopelias; it lifts and spreads the tail the same way in 

 cooing. The geopelias show in their juvenal plumage their relation to the turtle-dove. The Inca- 

 dove is probably the lowest; G. cuneata is the highest; G. humeralis is probably a branch off by itself, 

 but standing nearly as high as cuneata. 



"The bronze-wings are somewhat related to the geopelias, as is shown by their behavior, but they 

 have retained most of the turtle-dove pattern, 



"There are wild species so much alike that an unpracticed observer could see no difference in 

 appearance. Yet the voice and general behavior are very different. The specificity seems to per- 

 meate to every element of the bird's constitution. 



