64 COUNT T. SALVADORI ON MR. ELLIOT's [Jan. 14, 



38. Ptilopus porphyreus. 



This is not a Moluccan nor a Papuan species ; still I may mention 

 that, to avoid the confusion with Columba porphyracea, Temm. ( 1 822), 

 it would be better to call it P. roseicollis, Wagl. Besides Java it 

 inhabits also Sumatra, as has been stated by Bonaparte. I have seen 

 many skins from Sumatra, collected by Dr. Beccari. 



42. Ptilopus ornatus. 



The authority of Laglaize for this species being found on Mount 

 Arfak is not correct, as JNIr. Laglaize was never there. Mr. Laglaize's 

 specimens, which I have seen, are from Amberbaki, a locality far away 

 from Mount Arfak. 



44. Ptilopus perlatus. 



The locality Aru Islands does not belong to this species, but to 

 P. zonurns. The two are representative forms, one living in Northern 

 New Guinea, Jobie, and Salwatty, and the other in the Aru Islands 

 and in the south of New Guinea, on the Fly River, where D'Alber- 

 tis has lately collected several specimens entirely agreeing with those 

 from the Aru Islands. 



45. Ptilopus zonurus. 



Mr. Elliot could have added many quotations to the synonymy of 

 this species ; all the references to P. perlatus from the Aru Islands 

 belong to it. 



Beside the type, I have seen many specimens of this form from 

 the Aru Islands and from the Fly River ; and all of them show the 

 grey band at the tip of the upper surface of the tail. P. sonurus 

 has in that respect the same relation to P. perlatus that P. iozonus 

 has to P. jobiensis. It is important to notice that P. perlatus and 

 P. jobiensis are respectively the northern representative forms of P. 

 zonurus and P. iozonus. Mr. Elliot's statement that P. zonurus is 

 barely distinguishable from P. perlatus is rather inconsistent, after 

 he has accepted as distinct P. jobiensis and P. iozonus, which differ 

 exactly in the same particulars as P. zonurus from P. perlatus. 



47. Ptilopus pectoralis. 



The synonymy of this species is not correct. Instead of Columba 

 virens, Less. Yoy. Coq. descr. $ [««c], it ought to be Columba 

 cyanovirens, Less. Voy. Coq. Zool. i. 2, p. 713 (1828). The name 

 of C. cya7iovirens was given to the female of P. superbus and to the 

 present species ! It is important to notice the mistake, as, if Lesson 

 had really named this species C. virens, this name would have had 

 priority over that of C. pectoralis, Wagl. Isis, 1829, p. 739. From 

 the localities Mr. Elliot has left out Koffiao {Beccari). 



48. Ptilopus viridis. 



S. Miiller and many others after him have said that this species is 

 also found in New Guinea, near Lobo. But this is a mistake which has 

 arisen from S. Miiller having (Verb. Land- en Volkenk. p. 22) given 

 the name of Columba viridis to a young specimen of P. pectoralis. 



