( 120 ) 



blue on the head, and that there is a tendenc}- in them for a greater extent of the 

 hlne on the crown : and the Snmba specimens have mostly, though not always, a 

 distinct green line under the eye as well as above it, while in the Timor birds the 

 green line luuler tlie eye is not visible. In addition to all (iiis, I find that the beak 

 of the Snmba birds is larger, exceeding that of the Timor birds about 2 to 3 mm. in 

 the length of the cnlmeu, and about 1 to 3 in height, the same sexes being 

 compared, which is necessary, because the malcH have generally slightly larger bills 

 than the females. The wing also of the Sumba form is generally longer. The wing 

 measures in Timor specimens from 144 to lo4 mm., the average being aliout 1.50 to 

 152, only one out of seventeen having the wing fully 154 mm. long. In Snmba 

 skins the wing measures 150 to 156 mm., the majority of the males having it 155 

 to 156 long, none reaching below 147 mm., a length found only in one female. 

 Under the circumstances it becomes advisable to separate the Snmba form as 



Trichoglossus haematodus fortis snbsp. nov. 



Typus (? ad. Waingapo, Snmba, A. Everett coll., September 1896. 



50. Ptistes jonquillaceus (Vieill.). 



cJ. " Iris orange ; cere yellowish brown ; bill orange-red, witli yellow tiji, 

 mandible reddish orange : feet greyish black, claws black." ? . " Iris deep brownisli 

 orange ; cere light yellowish brown." A fine series, shot in .Inly at Atapupu. The 

 wings of the 7nales measure 180 to 190 mm. Most of the specimens were in 

 moult in Jnly. 



57. Cacatua parvula (Bp.). 



Two skins from Atapn))n. From Schlegel {^fl/s. P. B., P.vftnri, p. 137), 

 Salvador! {C<it. B. Brit. Mas. XX. p. 120, footnote), and Finscli we learn tliat the 

 tyj)e of C. parvula came from Semao, near Timor, and that it agrees with Timorese 

 skins. Our two Timor skins may therefore be taken as typical C. parvula. In 

 comparing them with our series of a dozen skins from Flores and Lombok, I find 

 that all of them have conspicuously longer, thicker, and stronger beaks tlian the 

 Timor birds. This observation is not new, for it had been made by Sclilegel (/.r.) 

 and by Finsch {Papageien I. p. 300). Indeed, tlie strong beak was the reason 

 why Finsch considered the Flores and Lombok specimens to belong to the equally 

 hirge-billed C. snlpharea of Celebes, his oliservation that they Jiacl, like €. sulplmrert., 

 a yellow patch on the ear-coverts, while this was wanting in the Timor birds, being 

 erroneous, for all the Lombok, Sambawa, Flores, and Timor birds have the ear- 

 coverts pale yellow, while it is of a much darker orange-yellow in C. sulphurea from 

 Celebes. That is also the reason why Salvador! says that most, but not all, C iiarruhi 

 had less strong bills than C. suljihircn. With the series in the Rothschild Museum 

 and others which I have seen, and with the confirmatory observations of these authors, 

 there can be no longer any doubt that the so-called " variation " in the size of the 

 bills of C. parvula is not " individual," but local. I therefore propose the name of 



C. parvula occidentalis snbsp. nov. 



for the large-billed form from T,ombok and Flores (type : i ad. Lombok, .Inly 1896, 

 A. Everett coll.), tliat from Timor and Seniao rcniaining C. parvula parvula. 



