878 Birds of Celebes: Anatidae. 



no mention of the occurrence there of such a bird as the male N. castaneum in 

 his "History of the Birds of New Zealand". 



Sumba was recorded for Mareca punctata (:= N. castaneum) by Meyer 

 (Verh. zool.-bot. Ges. Wien 1881, 767). The specimen certainly is not identical 

 with two males of N. castaneum from Australia, being very much smaller, besides 

 being in the female type of dress; it is smaller in the bill (probably therefore 

 a female) and a little greyer than our six specimens of the typical Nettion 

 gibberifrons from Celebes, but otherwise we find it to be identical therewith. A 

 specimen from Aru (C 6293: Riedel — a new locality) is similar to the Sumba 

 example. 



The locality Java was indicated for Anas punctata (= castanea) long ago by 

 Lesson (Tr. d'Orn. 1831, 252). The author of the species was unknown to him ; 

 he suggests with a "?" Horsfield, but this may have been due to the locality 

 Java being perhaps indicated on the label, this island having been then recently 

 made known in the ornithological sense by Horsfield's researches. No one 

 versed in the labelling of those days will attach much importance to the loca- 

 lity. On the other hand N. gibberifrons has been recorded and described from 

 Java by Dr. Vorderman; there is also an example from Java in the Dresden 

 Museum similar to Celebesian specimens of N. gibberifrons, and one in the British 

 Museum. 



From Celebes Anas punctata (■= castanea) was recorded by Prof. Reichenow 

 (J. f. O. 1877, 218; 1883, 122) and this remained the only notice of its having 

 occurred there. ') The probability of its having been N. gibberifrons has been 

 already suggested by Prof. W. Blasius, and Prof. Reichenow, in answer to 

 an inquiry by ourselves, kindly informs us that "the specimen in question is a 

 female and may therefore belong to gibboifrons, if that species is distinct'). 



Celebesian specimens of N. gibberifrons differ from Australian females of 

 iV. castaneutn (if rightly determined) by the small size of the former and 

 the cinnamon tint of the under surface, as compared with the buff whitish 

 of N. castaneum — the middles of the feathers being, of course, brown. In 

 New Zealand N. gibberifrons is larger than in Celebes, and no one has as yet been 

 able to show how it is to be distinguished from the female of N. castaneum. 

 The name gracilis was proposed for it by Buller, but afterwards withdrawn^). 



1) The "Anas punctata Tem." of Tiiisch (Neu Guinea 1865, 183) is placed by Salvadori in the 

 synonymy of Dendrocycna guttata, apparently with perfect right. 



2) Prof Eeichenow suggests that the males in the dress of gibberifrons in the Zoological Gardens of 

 London and elsewhere may have been the younger examples of castaneum, not yet in full plumage. But the 

 pair which bred in the Regent's Park were at least 3 years old in 1882. 



'■'} The extraordinary dissimilarity of the male in breeding plumage of N. castaneum promises to afford 

 food for philosophical investigation. At present it appears not improbable that the nuptial plumage was not 

 acquired by a very gradual process, for then we might expect to see lower stages of this dress in the male 

 of N. gibberifrons. It really looks as if the male of N. castaneum must have come into its nuptial dress aU 

 at once, just as under domestication an abnormal individual is sometimes bom which ''breeds true". 



