770 BULLETIN 50, UNITED STATES NATIONAL MUSEUM. 



of middle toe coherent with outer toe (or else, in genus PJmnicircus "■ 

 only, with inner side of tarsus feathered). 



The rano;e of variation in external features is so great in the group 

 known as Cotingidne that it is very difficult to discover external 

 characters additional to-those given above, which are common to all. 

 Doctor Sclater ^ separates the family from the Tyrannida; and Pipridse 

 by the alleged pycnaspidean tarsi ; but, while the tarsal envelope is 

 never exaspidean, as in the two groups mentioned, it is by no means 

 always pycnaspidean, this type of structure being typically developed 

 in less than half the genera, the majority having the tarsal envelope 

 taxaspidean or modified holaspidean.'^ 



Owino; to lack of sufficient knowledge of the internal structure of 

 the various genera, the limits of the family, as well as its relationship 

 to allied groups, are at present uncertain, and those here assigned 

 are merely tentative. If, however, the Pipridse are to be recognized 

 as a separate family, it becomes necessary to annex to the latter 

 the germs Laniocera (usually considered a member of the Cotingidse), 

 on account of its typically exaspidean tarsi and Piprine foot-structure, 

 to transfer from the Piprid^e to the Cotingidse the genus Lanii- 

 soma, and to eliminate, as a separate group, the genus Rupicola. 

 Whether Phcenicircus should be left with the Cotingidse, associated 

 with Rupicola, or placed by itself, I am not able to determine. At 

 the same time, after having defined the groups Cotingidse and Tyran- 

 nida^ by the difference in their tarsal envelope (apparently the only 

 available external character, while our knowledge of their internal 

 structure is much too fragmentary to be of service in the matter), 

 it becomes necessary to transfer from the latter to the former certain 

 forms which certainly belong there unless other genera to which they 

 are evidently related are wrongly placed in that group. These are 

 the genera Sirystes, allied to Lipaugus and Casiornis; and three 

 species removed from the Tyrannine genera, Myiarchus, Elsenia, and 



« Phoenicircus is so different from the true Cotingidse in its foot-structure that its 

 proper allocation is doubtful. It may possibly be more nearly related to Rupicola, 

 as suggested by Sclater and others.- 



b Cat. Birds Brit. Mus., xiv, 1888, 2. 



c The typical, or nearly typical, pycnaspidean tarsus is possessed only by Cephal- 

 opterua, Pyroderus, Querida, Hseviatodcrus, Calviceps, Gijvincdenis, Procnias, Tityra, 

 Lathria, Carpodedes, Xipholena, Cotinga, Calyptura, and Idiotriccus, the last named 

 (hitherto referred to the Tyrannidse) having not only the planta tarsi but also the 

 lower portion of the acrotarsium covered with small, roughened, almost tuberculous, 

 scutella. In Erator, Platypsaris, Pachyrhamphus, Xenopsaris, Ampelion, Phihalura, and 

 Heliochera there is a continuous definite row of rather large scutella along each side of 

 the planta tarsi, the structure being therefore essentially if not typically taxaspidean. 

 In Lipaugus, Casiornis, Sirystes, Tolmarchus, and Attila there is a single row, either 

 along the median line or occupying the outer and part of the inner side, while in 

 Euchlornis, Stictornis, and lodopleura the whole planta tarsi is without any sort of 

 scutellation. 



