JTodd-Carriker : Birds of Santa Marta Region^ Colombia. 295 



lationships of the various forms of this genus are still involved in 

 much obscurity, we are inclined on the whole to keep the present form 

 specifically distinct. The series at hand shows considerable variation 

 affecting the amount of spotting on the throat and breast. 



This most interesting bird belongs to the Paramo Zone, having been 

 found by Mr. Brown on the Paramo de Chiruqua at 15,000 feet, and 

 by the writer in the same general region, from 12,000 to 16,000 feet. 

 It is found only where there is water, as for example along the streams 

 coming down from the lakes and melting snow. It always occurs in 

 pairs, and is not particularly shy. It keeps to the rocks in and along 

 the edge of the water, and evidently feeds entirdy on aquatic insects. 



248. Fumarius leucopus agnatus Sclater and Salvin. 



Furnarius agnatus Sclater and Salvin, Nom. Avium Neotrop., 1873, 61, 159 

 ("Santa Marta"; orig. descr. ; type now in coll. Brit. Mus.). — Salvin 

 and Godman, Ibis, 1879, 197, foot-note, and 1880, 170 (Valle de Upar). — 

 VON Pelzeln, Ibis, 1881, 40S, part (Valle de Upar and " Sierra Nevada 

 de Santa Marta"; descr.; references). — Sclatek, . Cat. Birds Brit. Mus., 

 XV, 1890, 14, part (Valle de Upar; descr.). — Bangs, Auk, XVI, 1899, 137, 

 in text ("Santa Marta"). — Shaepe, Hand-List Birds, III, 1901, 47, part 

 (range). — Beabourne and Chubb, Birds S. Am., I, 1912, 222, part (ref. 

 orig. descr.; range). 



Furnarius leucopus var. agnata Dubois, Syn. Avium, I, 1900, 205, part (ref. 

 orig. descr.; range). 



One specimen : Rio Hacha. 



Furnarius agnatus was described from two specimens secured by 

 G. Joad in 1870, at Valle de Upar and Santa Marta respectively. In 

 the Catalogue of the Birds in the British Museum these two examples 

 are designated as.cotypes, but Salvin and Godman distinctly state that 

 the type came from Valle de Upar, and the measurements quoted in the 

 original description, and later by von Pelzeln, apparently bear this out. 

 The question of the type-locality becomes very important in view of 

 the discovery that the birds inhabiting the Goajira Peninsula, and ex- 

 tending eastward into the arid coast region of Venezuela, are subspe- 

 cifically separable from those of the restricted Santa Marta region 

 and Magdalena Valley. Our single Rio Hacha specimen (No. 45,637) 

 is very pale throughout, with the cinnamon rufous shading below much 

 paler and more restricted; the wing is longer also,' and the bill notice- 

 ably longer (26 mm.). The measurements correspond well with those 

 given by von Pelzeln for the type, and also in general with those given 



