G98 llecently published Ornithological Works. 



and two others liave eacli cue subspecies, making 22 in all. 

 Four of these are here di'^criminated for the first time. 



For the species usually known as Chamepelia craziana 

 Prev. & Knip, a new genus Eupelia is provided. As regards 

 C. pnsserina, a Liunean species and the type of the genus, 

 a further change is proposed, and it is worth wliile perhaps 

 to follow the details of this in order to show the difficuliies 

 of the modern taxonomist. 



Linucieus in describing his Columha passerijia in 1758 gives 

 a diagnosis which will suit the Ground-Dove o!' the southern 

 United States equally well with that of Jamaica and 

 Barbados, all three of which are now regarded as distinct 

 subspecies. Linnaeus based his diagnosis on Sloane, Catcsby, 

 and Willughby^s accounts, whose respective localities were 

 Jamaica, S. Carolina, and Barbados. 



Mr. F. M. Chapman in 1892 discussed this matter and 

 came to the conclusion that Jamaica should be accepted as 

 the type locality of Linnaeus' species as Sloane's reference 

 was the first, and proposed the name " ten-estris " for the 

 continental form. Mr. Todd has now discovered that Bona- 

 parte, in a review of the genus published in 1855, distinctly 

 restricted the application of the name '^ jjcisserina " to the 

 continental form figured by Catesby, and claims that this 

 form should be known as C. passerina passerina and not 

 C. passerina terrestris as proposed by Chapman, while for 

 the Jamaican and Barbadian subspecies the names C pas- 

 serina jamaicensis (Maynard) and C. passerina antillarum 

 Lowe must be substituted. 



The ultimate decision of this question must rest we 

 presume with the "checklist'' committee of the A.O.U. 

 or with the nomenclature committee of the International 

 Zoological ('ongrcss. So far <is we are aware, there is no 

 definite rule laid down in the code dealing with a case of 

 tliis sort. It seems, however, a sai'er method to follow 

 Chapman's method, as this leads move surely to a final 

 decision, rather than Mr. Todd's, \vhose conclusions may be 

 presently upset by the disco\ery of a reviser pi'evions to 

 liunapartCo 



