284 PROCEEDINGS OF THE NATIONAL MUSEUM vol. los 



L. I. smythiesi 



Sarawak: Wing, 50, 50, 50.5, 50.5, 50.5, 51, 51, 52, 52, 52, 52. Tail, 35, 36, 

 37, 37, 37.5, 38. B. I., 19, 20, 20, 20.5, 20.5, 20.5, 21, 21. 

 L. I. leucogastra 



Siam: Wing, 49, 49, 49.5. Tail, 33.5, 34, 37. B. I., 17.5, 17.5, 17.5. 



Malay Peninsula: Wing, 48.5, 50, 50, 50.5, 50.5, 50.5. Tail, 34.5, 35, 35+, 



35.5, 35.5, 36.5. B. I., 17, 17.5, 18, 18, 18. 

 Sumatra: Wing, 49.5, 52.5. Tail, 35.5, 37. B. I., 17.5, 17.5. 



Lonchura punctulata (Linnaeus) 



Linnaeus (1758, p. 173) originally described the nutmeg mannikin 

 as Loxia punctulata, listing its range simply as "Asia." His descrip- 

 tion was based entirely on the description and plate of Edwards 

 (1743, p. 40). In the 12th edition of the "Systema Naturae" (1766), 

 Linnaeus added a reference to Brisson, who specifically mentioned 

 Java. Kloss (1931, p. 364) considered that Linnaeus thus acted as 

 his own "first reviser," and that the type locality of Loxia punctulata 

 should be restricted to Java. Hellmayr (1914, p. 60) had earlier 

 come to the same conclusion. This is at variance with the treatment 

 of Baker (1926, p. 91; 1930, p. 223) and other authors who have used 

 the Linnaean name for the Indian rather than the Javanese race. 



The identification of Linnaeus' name punctulata with a particular 

 population of the nutmeg mannikin must rest entu'ely on Edwards' 

 description and plate, since this was the sole reference in the 10th 

 edition of the "Systema Naturae." If this description and plate are 

 identifiable as one of the subspecies of Lonchura punctulata, any sub- 

 sequent restriction by Linnaeus or anybody else is superfluous. 

 Examination of Edwards' plate and accompanying description make 

 it evident that it cannot possibly apply to the Javanese form, but 

 must be that of India, as advocated by Baker. The evidence is as 

 follows : 



1. "The bill is of the shape and bigness of our Green-finch's bill . . ." The 

 plate clearly shows a large-billed form of L. punctulata; the Indian bird is large- 

 billed, the Javanese small-billed. Culmen length of Chloris chloris (England), 

 13 mm.; Indian Lonchura punctulata, 11.5 mm.; Javanese L. punctulata, 10 mm. 



2. Although too much reliance should not be placed in Edwards' statement 

 that the bird is figured in its natural size, still the wing-length of the figure matches 

 Indian specimens exactly, and is far larger than Javanese specimens. 



3. "Head, neck, beginning of the breast, back, wings and tail, are of a dark 

 reddish brown ..." True of all except the tail of the Indian race; the Javanese 

 bird has the dorsal surface dull grayish brown, not dark reddish brown. 



4. As for the tail, which is described and figured as "dark reddish brown," 

 this does not match any form of the species, and is certainly not the Javanese 

 bird in which the tail is distinctly ashy gray. The shape of the tail in the plate 

 is also inaccurate, unless the specimen portrayed lacked the long pointed central 

 rectrices, which are yellow in the Indian bird. A specimen of the Indian race 

 in which the bright yellow, pointed central rectrices were missing, and in which 



