514 Annals of the Carnegie Museum. 



of feathers on the sides of the tarsus, most prominent on the outer 

 side and towards the heel. In C. buckleyi the tenth (outermost) 

 primary is very nearly as long as the next three, instead of decidedly 

 shorter. C. talpacoti and C. buckleyi agree with each other in having 

 the wings uniform dusky, all the other species having the remiges more 

 or less rufous. C. passerina differs from the other species in its 

 prominently squamate head and breast, but this character is more or 

 less evident in the juvenal stage of the remaining species, indicating 

 their derivation from a common ancestral type. 



Range. — Tropical America in general, north regularly to the Ber- 

 mudas, South Carolina, Texas, and Arizona; south to Paraguay and 

 northern Argentina. 



Taxonomic history. — Following Linnaeus, most of the earlier writers 

 referred the species of this genus — of which C. passerina was the 

 earliest and best known — to Columba of that author. Stephens, 

 however, placed it in his genus Goura, and in 1825 Spix described two 

 of the species under the generic heading Columbina, for which he 

 gave no diagnosis and fixed no type. The following year Boie pro- 

 posed the genus Columbigallina, with Columba passerina Linnaeus as 

 type. Passing by Spix's name, Columbigallina was taken up by the 

 American Ornithologists' Union in 1886 as the earliest name for this 

 genus, and not until 1907 was it discovered to be antedated by Columbi- 

 gallina Oken, 1817. This "long, badly constructed name, without 

 generic characters," however, met with but little favor outside of 

 America, most European authorities falling back on Chcemepelia 

 Swainson, 1827. This author cited as "examples" of his new genus 

 Columba passerina Linnaeus and Columba squamosa Temminck, but 

 designated no type, and it is further to be noted that one of the 

 characters given in his diagnosis, "the sides of the tarsi feathered," 

 does not apply to either of the above species. Because of this dis- 

 crepancy Reichenbach proposed in 1862 to restrict Swainson's name 

 to the species with feathered tarsi (for which Bonaparte had mean- 

 while established a genus of his own, Talpacotia), and to place C. 

 passerina and its allies in a new genus, Pyrgitocnas, but such action 

 was of course clearly invalid according to present rules. In 1841 

 Gray fixed the type of Chcemepelia 2 as Columba passerina Linnaeus. 



2 It should be noted, however, that in this instance Gray used the name in its 

 corrected form, Chamcepelia. Should this emendation be considered as a distinct 

 name the fixation of the type of Chcemepelia would have to date from J. E. Gray, 

 1855- 



