ART. 4 BIRDS OF THE MENTAWI GROUP RILEY 19 



Collocalia Unchi linchi from Java, of which the Museum now has 

 a good series, is so very different in color from the races credited to 

 this species that it is very doubtful if they should continue to be 

 regarded as forms of it. The Javan series stand out from the other 

 races of the species in having the upper parts washed with dark ivy 

 green ; all the other races of the species, at least so far as represented 

 in the United States National Museum, have the upper parts washed 

 with dull blue-green black. These colors are hard to match or define, 

 but perfectly obvious to the eye. The differences are so pronounced 

 that it would be better, from a practical standpoint, to recognize 

 C. I. linchi of Java as a distinct species without subspecies. 



HEMIPROCNE COMATA COMATA (Temminck) 



Cijpselus comatus Temminck, PI. Col., livr. 45, April, 1824, pi. 268 (Su- 

 matra). — Chasen and Kloss, Ibis, 1926, p. 283. 



Five males and four females, Siberut. 



Apparently not different from Sumatran specimens. The National 

 Museum also has it from Nias, Tana Bala (Batu Islands), and North 

 Pagi Island. 



HIEROCOCCYX FUGAX (Horsfield) 



Cueithis fugax Horsfield, Trans. Linn. See. London, vol. 13, 1821, p. 178 



( Java ) . 

 Hierococcyx fugax Chasen and Kloss, Ibis, 1926, p. 283. 



Two immature birds, Siberut. These are, as Chasen and Kloss 

 remark, " too young to deal with subspecifically." 



CACOMANTIS MERULINUS THRENODES Cabanis and Heine 



Cacomantis threnodes Cabanis and Heine, Mus. Hein., vol. 4, 1862, p. 19 



(Malacca). 

 Cacomantis merulmus threnodes Chasen and Kloss, Ibis, 1926, p. 284. 



Four males and two females, Siberut. 



Only two of the above are fully adult. The United States National 

 Museum contains only one specimen from Malacca; it is deeper in 

 color on the breast and belly than the two adult males from Siberut, 

 but Chasen and Kloss -" have compared Federated Malay States speci- 

 mens with those from Siberut and found them not separable. 



Oberholser ^^ diagnosed the Nias Island form as Cacomantis Tneru- 

 linus subpallldus, assigning as characters smaller size with head and 

 lower parts paler. The character of smaller size does not hold, but 

 it is paler than the specimens from Siberut. Whether the differences 

 are great enough to warrant recognition I am unable to dicide with- 

 out examining additional material. It was founded on two males, 



"Ibis, 1926, p. 284. 



=8 Smiths. Misc. Coll., vol. 60, no. 7, 1912, p. 4. 



