SYLVICOLID^. 161 



and notch in both mandibles separates it from such of the Vireonidse 

 as have nine primaries. To the Tanagridse, through the slender- 

 billed forms as Ghlorospingus, Nemosia, Ghlorochrysa, etc., the 

 relationship is very close ; so much so that, by many, both families 

 are included in one. What the real differences are, I may hereafter 

 be able to point out more satisfactorily than I can at present. 



The American Motacillidse are distinguished by the emargination 

 of the outer, and the great elongation of the inner secondaries, as well 

 as by other features referred to under that family. Anthus, in par- 

 ticular, differs in the lengthened and slightly curved. hind claw. 



There is, perhaps, no family to which the relationship is closer 

 than to the Gserebidse. Of equally small size, and, to some extent, of 

 ■ a somewhat similar style of coloration, it is not to be wondered at 

 that many species in each family have been indifferently assigned to 

 either. The genus Jlelminthophaga, for instance, can scarcely be 

 so defined as to distinguish it from Conirostrum, excepting by the 

 characters of the tongue, so rarely preserved in a skin. What the 

 external features of distinction are, I hope to show hereafter. I am 

 by no means sure that some species even now retained among the 

 SylvicoUdse would not be more appropriately placed in Gserebidse, 

 as Eelminthophaga bachmani, Parula gutturalis, etc. 



The tongue in the SylvicoUdse is horny for the greater portion of 

 its extent ; more or less deeply bifid at the tip for about one-fourth 

 or one-fifth the length, the branches fringed or lacerated along their 

 external margin. It is short and rather broad at the base, and not as 

 extensible, as in the Gserebidse. The essential difference in structure 

 from that of the Gserebidse seems to be that, in some of the latter, 

 as Olossiptila and Gerthiola, there is a second vertical plane erected 

 along the inner edge of the bifurcation or division of the tip of the 

 tongue, and more or less perpendicular to it, which is itself lacerated 

 or fringed, so as to increase materially the size of the terminal brush. 

 In Chlorophanes and Dacnis this vertical plane is folded outward 

 upon the horizontal lamina, and perhaps partially or entirely ad- 

 herent, and thickening considerably the inner portion of the fork. 

 The primary bifurcation of the tongue, however, in all the Gserebidse, 

 is also much deeper (about one-third the whole length), and the 

 lateral fringe extends much further along the base. There are other 

 differences in the tongues of the Gserebidse, of generic import, which 

 will hereafter be dwelt on more at length. 



To the general character of the tongue in the SylvicoUdse, however, 

 that of "Dendroica tigrina" forms a striking exception in its approxi- 

 mation to the Cserebine character, especially that of Gerthiola. The 

 11 IToTember, 1864. 



