THREE SPECIES OF LONCHURA — PARKES 287 



L. J), catervaria Koelz (1954, p. 19), described from the Khasia Hills, 

 represents a population intermediate between punctulata and sub- 

 undulata, and is best placed in the synonymy of the latter. It will be 

 remembered that subundulata itself represents a rather variable inter- 

 mediate population. 



Specimens examined: 



L. p. yunnanensis: Yunnan: Vicinity of Tengyueh, 23; Nantien, 1; Shweli- 

 Salaween Divide, 6. Burma: Myitkina, 1; 6 miles north of Myitkina, 1; 

 Singhaling Hkamti, 1; Washaung, 1; Sinlumkaba, 1. 



L. p. yunnanenszs X iopeZa intermediates: Yunnan: Mengtz, 4. 



L. p. topela: Taiwan, 40; Hainan, 23; Fukien, 6; Kwangtung, 1; Tonkin, 5; An- 

 nam, 2. 



L, p. topela X subundulata intermediates: Burma: Pegu, 2. 



L. p. subundulata: Assam, 5; Khasia Hills ("catervaria"), 10. 



Also series of other races in USNM and AMNH. 



Lonchura malacca (Linnaeus) 



Delacour (1943, p. 83) united three groups of forms which had often 

 been considered separate species : L. malacca, L. jerruginosa, and L. 

 atricapilla. He used the nsune Jerruginosa for the combined species, 

 overlooking the fact that Loxia malacca Linnaeus, 1766, antedates 

 Loxia Jerruginosa Sparrmann, 1789. The desirability of merging the 

 three groups has been questioned by some authors, but I am not 

 prepared to go into this question here. The present paper is con- 

 cerned only with that portion of the combined species formerly known 

 as Lonchura atricapilla, all forms of which lack white on the head 

 and underparts. 



It is tempting to consider L. grandis of New Guinea as part of this 

 species, since the adults closely resemble the atricapilla group. Dela- 

 cour (1943: p. 83) considered grandis and jerruginosa (=malacca) to 

 comprise a superspecies, giving the large bUl and extensively black 

 underparts of grandis as specific characters. Even more important, 

 in my opinion, is the pattern of the juvenal plumage, in which grandis 

 differs markedly from all races of malacca in being streaked on the 

 throat. 



Salomonsen (1953, p. 265) pointed out that a revision of this species 

 was badly needed, and he reviewed it briefly himself. The large 

 area ascribed by Salomonsen and other authors to the race atricapilla 

 involves a number of well-marked clines, which can be divided into 

 several subspecies and intergrades. Salomonsen's treatment of the 

 Philippine populations is critically discussed (p. 290) after a review 

 of the mainland forms. 



The distribution of L. m. sinensis is substantially as given by 

 Salomonsen (1953, p. 266), namely Malay States, southern Siam, 

 and lowlands of Sumatra. The diagnostic characters can, how- 



