102 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



yellow instead of pinkish red, and the deep red of the crown is very 

 abruptly defined anteriorly against the whitish of the forehead, instead 

 of blending- into the color of the frontlet, there being no trace ©f a whitish 

 frontal band in C. carolimis ; the abdomen in C. tricolor is bright scarlet, 

 instead of light pinkish red. In C. carolimis the tibias and lower tail- 

 coverts have a few longitudinal, irregularly sagittate marks of dusky, 

 whereas these parts in tricolor are heavily banded with transverse, 

 somewhat V-shaped bars. In carolimis only the inner webs of the middle 

 tail-feathers are barred or spotted with white, while in tricolor both 

 webs are so marked. C. tricolor is also very much darker colored un- 

 derneath than C. carolimis. The females of the two species do not re- 

 semble one another so closely as do the males. In that of G. tricolor 

 the red of the nape is both more restricted and duller than in that of 

 C. carolimis, the pileum is of a light smoky drab, instead of pure ash- 

 gray, and the forehead is dull whitish, tinged with yellow anteriorly, 

 instead of pinkish red. 



These comparisons apply to the most southern examples of C. carolimis 

 I have been able to examine, viz, a considerable number from Texas 

 and South Florida. 



In this species, as in most others, there is much individual variation. 

 In a majority of the adult males the red of the nape is continuous with 

 that of the crown, but in some (as in an example from New Granada, in 

 the collection of Salvin and Godman), the occiput is completely crossed' 

 by a Avide baud of light smoky drab, widely separating the orange-red 

 of the nape from the bright crimson vertical patch. Other examples 

 show a more or less complete coalescence of the two red areas, in varia- 

 ble degree, according to the individual. Such examples appear to in- 

 clude adults as well as young birds, so that age has apparently nothing 

 to do with the variation in question. In specimens having the crimson 

 of the crown widely separated from the more flame- colored red of the 

 nape there is a rather close resemblance to some specimens of C. lioff- 

 manni, in which, however, the abdomen is yellow instead of bright red, 

 the outer webs of the middle tail-feathers streaked, instead of barred or 

 spotted, and the nape decidedly red, instead of orange-yellow. 



3 (?). CENTUEUS EUBEIYENTEIS. 



Centurus ruhiventris, Swains. Anu. in Menag. 1838, 354 (liab. ignot.). — Gray, Geu. B. 



ii, 1849, 442.— Lawr. Ann. Lye. N. Y. ix, 1869, 206 (YucataD). 

 Zebrapicus rulrivcntris, Malii. Mon. Pic. ii, 1862, 248; iv, 1862, pi. cvii, fig. 1 ( c^ 



ad.). (Mexico?) 

 " Zebrapicus sivainsonii, Malh. 1845, in mns Britan." (Malherbe.) 

 " Ficus aurifrons ", Bonap. P. Z. S. 1837, 116 (nee Cousp. i, 1850, 119). {Fide Malherbe.) 

 Fkzebre a ventre sanguin, Malii. 1. c. 



Sal). — Yucatan. 



This bird, which appears to be a well-defined form, I have not seen, 

 and therefore copy Swaiuson's original description, and Mr. Lawrence's 

 remarks in Ann. Lye. N. Y., ix, 1869, pp. 206-7, which, with Malherbe's 



