Chapman — New Birds Peru, Bolivia, Brazil, Colombia. 267 



and breast streaked with black, the upper pails pronouncedly blacker and 

 the rump yellow, in strong contrast to the back. This female closely re- 

 sembles one of true uropygialis from Ecuador, but has the white terminal 

 markings of both wing-coverts and upper tail-coverts larger. 



Catamenia analoides griseiventris, new subspecies. 



Subspecific characters. — Male similar to male of Catamenia analoides 

 analoides (Lafr.), of the Peruvian coast region, but abdominal region 

 grayer, less white, the under parts therefore nearly uniform in color; 

 second to sixth primaries (from without) with less white on their outer 

 webs at the base; lower tail-coverts averaging paler and usually without 

 the buffy tips which are always present in true analoides. 



Type.— No. 129,178, Am. Mus. Nat. Hist., d 1 ad., Cuzco, Peru, Novem- 

 ber 16, 1914; H. Watkins. 



Specimens examined. — Catamenia analoides griseiventris. Peru: Type 

 locality, 1 d 1 , 1 9 ; Pisac (near Cuzco), 4 d\ 2 9 ; Chospyoc, 10,000 ft., 

 1 <?; Huaracondo Canon, 10,000 ft., 2 tf , 1 9 ; Torontoy, 8000 ft., 1 d 1 . 

 Ecuador: Valle de Cumbaza, Chimborazo, 3 d 1 . 



Catamenia analoides analoides. Peru: Lima, 2 d 1 . 1 9 ; Vitarte (near 

 Lima), 9 d", 5 9 ; Huacho (north of Lima), 1 d\ 2 9 ; Huaral (north of 

 Lima), 5cf,l 9 ; Sayun, 1 d 1 . 



Remarks. — Our excellent series of this species shows that most of the 

 differences between specimens from the coast and those from the table- 

 land of Peru, to which Taczanowski (Orn. Per., Ill, p. 21) long ago called 

 attention, are diagnostic. The fact that specimens from the tableland of 

 Ecuador agree with those from southern Peru, indicates that the race here 

 described occupies the puna or paramo in both Ecuador and Peru, while 

 true analoides is doubtless confined to the coastal region. 



It is interesting to observe that griseiventris, in its gray abdomen and 

 decreased amount of white at the base of the primaries, makes no approach 

 toward the Bolivian analis, which indeed more closely resembles the Lima 

 race than it does the one geographically nearest to it. 



The Colombian form. C. a. schistaceifrons, is also more like the one occu- 

 pying the Peruvian coast than the race which inhabits the intervening 

 mountains. 



