14 'bulletin 50, UNITED STATES NATIONAL MUSEUM. 



tiviitiucntl It may hv. and doubtless is, perfectly true that no more 

 than three l*as.seriform groups can be defined which will be equal in 

 taxonomic i-aidc to the families of other orders of birds; but the objection 

 to tliis meager allowance — and it is a very serious one — is that two of 

 the three groups contain together only about one-fifth the total num- 

 ber of species, so that there are still left about five thousand species in 

 the third. Ob\'iously. these fi^'e thousand species (more or less) must 

 be susceptible of segregation into a considerable number of more or 

 less trenchant groups: and there being so few grades of rank between 

 a family and a genus, what to call these groups becomes a ver}' serious 

 question. The ordinary terminolog}^ of zoology evidently will not 

 sufiice; and if no more than three families of Passeriformes are recog- 

 nized, a new and complicated nomenclature for the intermediate groups 

 becomes necessary. 



As a provisional expedient, 1 propose to call the Passerine "' families" 

 of (xadow '^S'ujjerfamilies,-- and retain the former term for such groups 

 of genera as can be trenchantly separated from all others. Whether 

 this action will necessitate a reduction or an increase in the number of 

 so-called families over that generalh' accepted can only be determined 

 after careful and thorough study of the entire order. This is a task 

 for Avhich the author of this work is unprepared, either as to time or 

 material. The best that he can do here is to limit investigation in this 

 direction to the American forms. Of course the result of such limited 

 research can not be entirely satisfactory; but it ma}^ serve to show, 

 perhaps more clearly than has been done before, which currently recog- 

 nized families can and which can not be characterized. Nothing is 

 more certain than that the commonh' accepted limits of some of the 

 so-called families of the Superfamily Oscines are purely artificial and 

 arbitrary. On the other hand, it is equalh' obvious that some groups 

 to which family rank seems due have been ignored or overlooked. 

 Until more is known concerning the internal structure of various forms 

 anv classification of the Oscines must be considered imperfect and 

 pro\isional. 



KKY TO THK St'BORDEKS OF J'AtSSERI FORMES. 



(I. Hallux \voak; feet syndactyle,' the deep iilantar tendons of Type I'- (desmopel- 

 niousV); cervical vertebrae 15; spina externa eterni long, simple. . .Desmodactyli. 



aa. Hallux the strongest toe; feet eleutherodactyle,Hhe deep plantar tendcHis of Type 

 VII^ (schizopehnousM; cervical vertebras 14; spina externa sterni short, 

 forked Eleutherodactyli. 



^In the syndactyle or desniopelmous foot the flexor perforans digitorum and flexor 

 halluciis loiifjus tendons are united at their crossing point by a vinculum. In the 

 eleutherodactyle or schizoi)elmous foot, on the other hand, these ten<lons are quite 

 separated from one another. 



-(l.\RKOD, Prbc. Zool. Soc. Lond., 1875, pp. 339-348; Gadow, in Bronn's Thier- 

 Reichs, Vijg., 1891, p. 195, ii, Systematischer Theil, 1893, pp. 224, 225; in Newton's 

 Dictionary of Birds, pp. 615-618. 



