302 Allen, Columbina vs. Choemejyelia. \fu^ 



type, and with only two species referred to it, namely, Cohi.mba 

 passeriTia Linn, and Columba squamosa Temm. Li 1841 Gray 

 (List Gen. Bds., 2d ed., 75) designated Columba passerina as its 

 type, a species he had in the preceding year made the type of Colum- 

 bina ! The other species was referred by Bonaparte in 1854 (Consp., 

 II, p. 85) to his new genus Scardafella, and later it became its type 

 by subsequent designation (Gray, 1855, Cat. Gen. and Subgen. 

 Bds., 100). Selby, m 1835 (Nat. Libr., Pigeons, 198), designated 

 "Columba Talpicoti Temm." as the type of Chcemspelia, and Swain- 

 son, m 1837 (Class. Bds., 349), cited the same species, under a differ- 

 ent name ("Columba cinnatnomina. Spix, II, [pi.] 75a, f. 1" = 

 talpacoti Temm.) as its 'example'; but both of these designations 

 were invalid, as the species selected was not originally mcluded in 

 the genus. Hence after squamx)sa was removed in 1854, passenna 

 was the only species left in the genus and it thus necessarily became 

 the type of Clicemepelia by restriction. But if Gray's act making 

 passenna the t}-pe of Columbina, in 1840, was valid, this would 

 render Chcemepelia a synonym of Columbina. 



The first step in the consideration of this question is to note the 

 fact that passerina was not nominally one of the four species origi- 

 nally referred to the genus Columbina, but Spix did include in it a 

 species — griseola — which is in reality only a slightly differentiated 

 subspecies of passerina. Of this, as will be showai later, there can 

 be no question. But the griseola of Bonaparte and of nearly all 

 subsequent authors was not the griseola of Spix. This explains why 

 "griseola" has been usually recognized as either a distinct species 

 or as a synonym of Columba minuta Linn., and renders it necessary 

 to consider the taxonomic history of not only Columbina griseola 

 Spix but also of Chcemepelia griseola Bonap. and of Columba mimda 

 Lmn. ^l^Fi 



Columba minuta Lmn. (Syst. Nat., ed. 12, 1776, 285) was based 

 exclusively on the Turtur parvus fuscus americanus Brisson (Om., 

 I, 1760, 116, pi. viii, fig. 2), which was poorly figured but exceedingly 

 well described, as is attested by the nilmgs of modem authorities 

 (see especially Salvadori, Brit. Mus. Cat. Bds., XXI, 1893, 481). 

 Bonaparte, who is responsible for much that is unfortmiate in ornitho- 

 logical nomenclature, was the first author to refer (Consp., II, 1854, 

 77, 78) C. minuta Lmn. to C. passerina Linn., as the young of the 



