( 371 ) 



little more white at the base of the dorsal feathers, but always much less than in 

 H. p. gi'iseiventris, of which, thanks to the kindness of my friend Dr. von Lorenz, 

 two of the types are now before me. In addition to the characters indicated in 

 the diagnosis of H. vidua, I find that the edge of the wing is cinereons (not pale 

 ferrnginons as in //. griseiventris), the chin whitish like the throat (not ferrugiuous), 

 and the upper surface of a much less reddish tint. 



Considering the apparent close relation between the females of H. grisekentris 

 and //. vidua, I at once suspected that the male of the former was most probably 

 of a similar style of coloration to those of //. poecilinota and allies, and might 

 have been mistaken by Pelzeln for the male of H. p. lepidonota, of which Natterer 

 collected a fine series at Marabitanas on the upper Rio Negro. Dr. Lorenz having 

 kindly sent me the whole series, I found among the skins an adult male collected near 

 Borba, where Natterer also obtained two of the type specimens of Pitbjs gri.sei- 

 tentris. This skin is marked by Natterer himself as the male of the latter — 

 a statement whi('h seems to have entirely escajied tlie notice of Pelzeln — and, as it 

 agrees with them in all structural details, there can he no doubt that it represents 

 the hitherto unknown male sex of P. griseiventris ! 



On comparing it with a large series of H. p. poecilinota and II. p. lepidonota, 

 the general resemblance is very close indeed, but the Borba bird differs from either 

 in having tlie upper tail-coverts pale grey, with only a narrow black subtermiual 

 bar just behind the white tip, while they are deep black tipped with white in 

 its allies. The base of the three median pairs of rectrices is distinctly washed 

 with cinereous (in H. p. poecilinota, H. p. lepidonota, and H. p). vidua they are 

 deep black), and the throat is decidedly paler grey than the breast, but not whitish 

 as in //. p. vidua. 



We have thus four geographical races of one t)'pe : the males not or very 

 slightly difl'erent, lint, the females with striking differences in coloration. The 

 most interesting fact, however, is that iu two of these forms — //. p. poecilinota and 

 H.jj. lepidonota — the females have attained the ornamental characters of the male 

 plumage, while in the two others — H. p. vidua and H. p. griseiventris — they are 

 still devoid of these, tlius apjiarently representing a more primitive state of 

 development. 



In tlie following lines I give the characters and the range of the four forms : — 



a. Hypocnemis poecilinota poecilinota Cab. 



Hypoenrmis poecilinota CabanLs, Arch. f. Naturg. 13. i. (1847) p. 212. tab. 4. fig. 2 (cJ) [British 

 Guiana]. 



Hab. British Guiana : Bartica Grove, Camacusa, Merume Mts., Roraima, 

 Rio Atapurow. Surinam : Maroni River. Cayenne : Rio Approuagne (Cherrie 

 coll., Mus. Tring). Venezuela : Mnnduapo and Nericagua on the Orinoco R. ; 

 Canra R. : Nicare, La Pricion, La Union, Snapure. 



d. Tliroat quite as dark grey as the rest of the underparts ; ujiper tail-coverts 

 black with white tips. 



Wing 65—71 ; tail 40—4.5 ; bill 17—18 mm. 



?. Forehead and sides of head, including chin, bright clear ferruginous, passing 

 mto rufesoeut brown on tiie pileum ; back olivaceous brown, and like the upper 

 wnig-coverts and inner secondaries with broad black subterminal and fulvous apical 

 cross-bands; upper tail-coverts olive brown witli fulvous tijis. Edge of the wing 

 pale ferruginous. Tail-feathers black, witli a. wiiite apical margin and a crramy 



