( 31 ) 



Xo. 2. Wing 115—121 mm. (The hept-marked form of all.) Sula Islands, and, it 

 is .said, Ceram (V). 



(). P. mehinocejih'ihi peiii>f/ensi.<i Hart, (nilinopus chrysoyrhoiis -pdingeiisis 

 Hart.. Nov. /OOL. 1898. p. Kio, Peling and Banggai). Entirely like No. 5, but wing 

 only 109—114 mm. Peling and Banggai. 



7. P. melanocephda x<iiithorvho<i (Salvad.) {lotreron xnnthorrhoa Salvad., Anii. 

 Mus. Civ. Gen. vii. p. (i71, 1875. Sanghir). Black occipital patch large, gular patch 

 very pale lemon-vellow, vent and shorter under tail-coverts deep orange, wing very 

 long, 130—139 mm. Sangi Lslands. 



8. P. mdanocephala tnlantemis subsp. nov. In every way like No. 7, but 

 smaller, wing 120—130 mm. Talaut Islands. 



Type No. 4444, S. l.irung, Talant Islands, ]\Iay 1897, collected by John 



Waterstradt's natives. 



The following specimens of P. m. aurescentior have been sent by Mr. Kiihn : 



3 (?, 2 ? ?, Kalidupa, January 1902. (Nos. 4567—4570.) 

 r. cJJ, Tomia, December 1901. (Xos. 4209, 4214, 4358-4301.) 

 3 J c?.. 1 .juv., 1 ? (y juv.). Binongka December 1901. (Nos. 4127,4128,4210, 



4212, 4213.) 



3 (?(?, 1 ?, Wantjee, December 1901. (Nos. 4211, 4436, 4437, 4400.) 

 2 cJcJ, 1 ?, Buton, November 1901. (Nos. 4158, 4161, 4170.) 



50. Carpophaga concinna Wall. 



Carpjopharjri condnna Wallace, 7^)(',s 1 865. p. 383 (" Matabello, Sanguir Island, 

 Aru [one small island west of] ; Banda Island, Ke Island (seen, but no specimens 

 obtained; Philippine Islands? [B. .AI.]." Typical locality Matabello! Cf. Cat. B. 

 Brit. Mus. xxi. p. 187.) 



Kalidupa (Nos. 4535, 4536, 4537). 



Binongka (Nos. 4188—4192). 



Tomia (Nos. 4326—4331). 



.\11 these are ty[iical light grey C. concinna, while on the Key Islands is a much 

 whiter form, ('. concinna separata Hart. I cannot in the least see the reason why 

 ^Messrs. ^[eyer and Wiglesworth (B. Celel>es ii. p. 617) should have taken the trouble 

 to warn me not to '' .split " C. concinna into subspecies. I am of oiiinion that I had 

 examined a sutficiently large material to know that there was no seasonal change in 

 the direction of my sepjarata, and that these birds did not migrate from Matabello to 

 the Key Islands. I have now, in the Tring Museum alone, 50 typical C. concinva 

 and 8 C. concinna separata for examination, and they show at a glance the differences 

 of the two races. Dr. Finsch (Xotcfs Leyilen Mus. xxii. p. 295) also mentions 

 " Uebergange," but I find my '' separata" a most distinct form, though of course 

 only a svibsjjecies ; so that there may be specimens " fast so gran wie Exemplare von 

 andern In.seln." confirming my view as to this subspecies. 



51. Carpophaga rosacea (Temm.). 



Cohimha rosacea Temminck, PL Col. 578 (1835 : Timor). 

 ? juv.. Binongka, 9. xii. IVIOI. " Iris dark crimson, feet pale crimson, 

 slate-grey." (No. 4193, Kiihn leg.) 



