318 Annals of the Carnegie Museum. 



those from the vicinity of Santa Marta, as well as with a series from 

 the Magdalena Valley and Rio Sinu, although there is, to be sure, a 

 great deal of variation. But Rio Hacha specimens, upon which the 

 subspecific name phainoleucus has been based, present certain peculi- 

 arities. In the males the black area of the throat and breast is more 

 restricted, leaving the throat superficially white in most examples; 

 there is more white on the forehead, sides of the pileum, and sides of 

 the head in general, so that these parts, instead of being black, freckled 

 with white, are white, speckled with black; the white external margins 

 of the primaries are more pronounced; the white tips to the rectrices 

 are perhaps a little larger ; and the upper parts in general average paler, 

 although there is considerable variation here. Were these characters 

 more constant it would be feasible to distinguish an eastern form, but 

 there are too many exceptions. Thus, a nlale in fresh plumage from 

 Lorica, Bolivar, Colombia (No. 52,678) is an exact counterpart of 

 the type of phainoleucus, and the examination of the series in the col- 

 lection of the American Museum of Natural History shows that there 

 is a great deal of variation. Females seem to offer better characters 

 for separation, and it may be that eventually it may be possible to 

 maintain phainoleucus on such grounds. All the Rio Hacha females, 

 together with two from Venezuela, are paler, more i)uffy, less rufes- 

 cent in general coloration, with the throat and sides of the head 

 nearly white, while the dusky streaking on the breast and lower throat 

 is nearly obsolete. But until a larger series is available we prefer to 

 ignore the supposed form. 



This ant-thrush is found sparingly over the whole of the semi-arid 

 lowlands from Santa Marta to Fundacion, being commonest along 

 the shores of the Cienaga Grande. It is perhaps more numerous in 

 the Rio Rancheria-Rio Cesar Valley, and at Rio Hacha it was fairly 

 common in the thick scrub. It is always found in open woodland or 

 scrub, never in the heavy forest. The birds are very tame, go about 

 in pairs, and are rather sluggish in their movements, 



269. Taraba transandeana granadensis (Cabanis). 



Two specimens : Tucurinca. 



This form, which Mr. Ridgway (without, however, having examined 

 specimens) doubtfully includes as a synonym of T. transandeana 

 transandeana,' is a well differentiated subspecies, as shown by a direct 



