, 26 , BIJLLETlEiN" 50, UNITKD STATES NATIONAL MUSEUM. 



natural and fairly well characterized group if the genera Acanthidops 

 and Sicalis be taken out' and Chloris and ChaunoproGtus added. ^ 

 Indeed these two groups (i. e., the " CoccothraustinsB " and "Fringil- 

 linse " as properly linaited) together eome very near being trenchantly 

 separated from all other Fringillidae; but the genus Passerina seems 

 to connect them, the latter being in all respects (as to external charac- 

 ters) except the shape of the bill like Leucosticte (a typical "fringil- 

 line" form), while the bill is very similar to that of a typical Eynbe- 

 riza. All of the forms comprising the above-mentioned groups are of 

 northern distribution (many of the genera being circumpolar), only the 

 genus Spiniis extending into the Neotropical Region proper, excepting 

 its near relation, Loximitris, confined to one of the Greater Antilles. 



These northern groups present little difficulty, except as to the 

 determination of the question whether they should be regarded as 

 constituting one large group distinct from all other Fringillidse or as 

 comprising a greater or less number of smaller grotips, of equal value 

 with similar groups in the larger assemblage of purely American 

 forms to be considered separately. Although -unable to fully satisfy 

 myself as to which course would be best, I have, for the present, con- 

 cluded to adopt the latter alternative; and, therefore, instead of recog' 

 nizing two groups, equivalent to Dr. Sharpe's Coccothraustinse and 

 Fringillinffi, as amended, or one group including the two, four groups, 

 Coccothraustese, Loxise, Pyrrhulse, and Fringillse, are provisionally 

 adopted. 



These coccothraustine and f ringilline types having thus been tempo- 

 rarily disposed of, there remains the very numerous assemblage of 

 peculiarljr American ^ genera. These, with the exception of the group 

 which 1 have here named Calcarieee (comprising the genera Passerhia, 

 C'alcarius, and RhyncJwphanes^ which are evidently related to Pala;- 

 arctic types *), are all peculiar to America (mostly to the Neotropical 

 Region) and with few exceptions not at all like any Old World types. 

 It is this group which presents the greatest difficulties in the way of 

 satisfactory classification. Not only do the different groups (or what 

 seem to be natural groups) run into one another in a most perplexing 



' Aavnthidops is certainly not a member of the Fringillinse, its nearest relation being 

 undoubtedly the "emberizine" genus Haplospiza. I am reasonably sure that S'caWs 

 also is an "emberizine" form (related to Haplospiza, Pseudochloris, etc.), notwith- 

 standing the remarkable superficial resemblance of some of the species to the truly 

 fringilline genus Si'rI.nus. 



'' Chaunoproctus seems to be a crass-billed Carpodamx, an approach to its characters 

 being seen in the insular Carpodaciis dmplus. 



'Excepting only Passerina and Calcarius, the latter chiefly American, since \:\o of 

 the three known species are peculiar to the Nearctic Region. 



' Whether there are terrestrial Fringilhe or specialized EmberiziB I am unable to 

 determine. 



