SYLVICOLID^. 165 



Icteria and Teretristis do not differ from the rest, althougli it has 

 been suggested to place the latter genus in Geerebidae. 



The Pabid^ {Parus septentrionalis, Auriparus Jiaviceps, Psal- 

 triparus minimus, Sitta canadensis) have the tongue thicker and 

 more fleshy than in Sylvicolidse ; only moderately horny at ends and 

 tips, and only slightly bifid and lacerated, much less than in Sylvico- 

 lidse, and about as in Turdidse, but less horny. 



The HiRUNDiNiD^ {S. horreorum and bicolor) have the tongue 

 short, broad, triangular, quite fleshy, with a shallow notch at end, the 

 sides scarcely or not at all lacerated. 



The ViREONiD.ai ( V. crassirostris and barbatula) have the tongue 

 more like the Paridae than the Sylvicolidse. 



The characteristics of Cmvx&tdm, as far as I have been able to 

 examine their tongues, I have already referred to in detail. The 

 other families of Oscines will hereafter be referred to. 



In the TyRANNiD.a) the tongues of Muscivora mexicana, Pitangus 

 derbianus, and Todirostrum are horny for most of their length, 

 nearly linear, or gently tapering to a blunt, almost truncated tip, 

 which is not bifid, but has several short incisions in the end. 



In the ToDiD.ai the tongue of Todus viridis is broad and linear 

 to the blunt tip (shaped like the bill), and throughout horny, thin- 

 edged and entire. 



In GALBULiD.a; a Odlbula from South America has the tongue 

 equally horny, but long, narrow, tapering gently to a point, and" 

 without any incision whatever. 



In the " Birds of North America" I have dwelt at much length 

 upon the characters by which the North American genera of Sylvico- 

 lidse are distinguished, and refer to that work for particulars, repro- 

 ducing here only some of the diagnostic tables, with a few modifica- 

 tions. These can doubtless be much improved ; but I have not time 

 at present to attempt to work them over again, and must content 

 myself here with dwelling in detail only on the forms of Middle and 

 South America. It will be noticed hereafter that while the Sylvico- 

 linse proper belong essentially to North America, it is in Middle and 

 South America that the Setophaginae have the greatest development. 



The following synopsis is an attempt at defining the higher sub- 

 divisions of the Sylvicolidse. In the large number of species, their 

 close relationships, and the very gradual transition from one form to 

 another, I have found it very difficult to make any arrangement by 



