192 PROCEEDINGS OF THE ACADEMY OF [April, 



Collocalia whitehead! Grant. 



Collocalia whiteheadi Grant, Ibis, 1895, p. 459. 



Chars, sp. — Similar in color to Collocalia loici, but dark shaft streaks 

 on lower parts less distinct ; averaging slightly larger, with unf eathered 

 tarsi, and usually more deeply emarginate tail. 



Wing, 127-140; tail, 50-57 ; exposed culmen, 5; tarsus, 11.5-13 mm. 



Type locality. — Monte Data, highlands of Lepanto, northern Luzon, 

 Philippine Islands. 



Geographical distribution. — Islands of Luzon and Palawan, Philippine 

 Islands. 



This very distinct species may be easily distinguished from all the 

 others with unfeathered tarsi, excepting Collocalia origenis, by its ^-ery 

 large size, and from that form by the well-marked color characters 

 already detailed. Specimens from Palawan are smaller than those 

 from Luzon, touching the minimum of measurements above given, 

 and also appear to be more brownish on the upper parts — differences 

 which, should they prove reasonably constant, would entitle the 

 Palawan bird to subspecific rank. 



Collocalia unicolor unicolor (Jerdon). 



Hirundo unicolor Jerdon, Madras Journ., XI, 1840, p. 238. 

 Cypselus concolor Blyth, Journ. As. Soc. Bengal, XI, pt. 2, 1842, p. 886 (nom. 

 nov. pro Hirundo unicolor Jerdon). 



Chars, subsp. — Resembling Collocalia ichiteheadi, but much smaller; 

 more brownish and more uniform above, the rump not appreciably 

 paler than the back, the pileum but little if any darker; lower surface 

 rather more brownish and more uniform, the throat usually not lighter 

 than the abdomen; lining of wing lighter. 



Wing, 112-120.5; tail, 50-56; exposed culmen, 4-4.5; tarsus, 9-10 

 ,mm. 



Type locality. — Coonoor Pass, Nilghiri Hills, southern India. 



Geographical cUstrihution. — Ceylon; and the western coast region of 

 southern India, north to Vengmia; western Himalayas. 



Although this species has commonly been either synonymized with 

 Collocalia fuciphaga or treated as a subspecies of it, the entirelj^ un- 

 feathered tarsi are a character fully sufficient for specific recognition; 

 and furthermore the upper surface is more brownish than in even the 

 least blackish forms of Colloc*lia fuciphaga. In fact Collocalia uni- 

 color is really more closely allied to Collocalia ichiteheadi than to Col- 

 localia fuciphaga! Dr. Hartert has given" as a reason for considering 

 C. unicolor a subspecies of C. fuciphaga that some Celebes birds similar 

 to C. unicolor in color, which he refers to C. fuciphaga, have no tarsal 



1 Ibis, 1896, p. 369 



