NOVITATES ZOOLOQICAK XXIII. 1916. 83 



cinnamon-rufous. Rump brown ! Wings, c? 145-148, sometimes 150-151, ? 136- 

 145 mm. 



Nile valley sooth to Wadi Haifa. All specimens I have seen from south of 

 Wadi Haifa were typical senegalensis, though some from Kerman in Nabia 

 (Hon. N. C. Rothschild coll.) and the Atbara were very pale. It would be inter- 

 esting to compare larger series from these districts. 



5. Streptopelia senegalensis cambayensis (Gm.) 



Culumba cambayensis G-melin, Syst. Nat. i. 2, p. 779 (1789 — "Habitat in Cambaja," i.e. on the 

 Bay of Cambay, North of Bombay. Ex Sonnerat and Latham). 



Differs from all the forms of S. senegalensis in the uniform dull earth-brown 

 npperside. Wing, S 125-132, ? 121-129 mm. 



Tropical India from the foot of the Himalaya to the Malabar coast, where it 

 is rare, to the east to the rivers Hoogli and Ganges (Baker). It is a mistake to 

 suppose that this form or ermannt extends to Asia Minor or Constantinople ! 



6. Streptopelia senegalensis ermanni (Bp.) 



Turtur ermanni Bonaparte, Compt. Rend. Acad. Paris 43, p. 942 (1856 — Bokhara. Types in Berlin 

 Museum. There is no doubt as to the form which Bonaparte described, though his description 

 is partially misleading, as he evidently had specimens of senegalensis, which he called cam- 

 bayensis, for comparison). 



Exactly like S. s. cambayensis, but larger: wings, <$ 135-145, ? 134-140 mm. 



This very closely allied form is not always separable, if only single specimens 

 — especially of doubtful sex — are compared. It inhabits parts of Turkestan 

 (Bokhara), parts of Persia, Baluchistan and Afghanistan, and Muscat in S.E. Arabia. 



There has been some uncertainty about the Yunnan form of Streptopelia 

 chinensis. Mr. Ingram called it Turtur ckinensis, but said : " These (i.e. a pair 

 from ' Mongtse ') agree with T. ckinensis fairly well, but have indications of dark 

 median streaks on the upper wing-coverts, and in one example the under tail- 

 coverts are almost white ; in fact, they appear intermediate between T. ckinensis 

 and tigrinus." Anderson identified specimens from Western Yunnan with 

 T. tigrinus, and also Bangs and Phillips called examples from Mengtsze 

 (" Mongtse ") Spilopelia tigrina, though also adding that they were " more or less 

 intermediate between tigrina and ckinensis." 



Comparing the Yunnan specimens I was astonished to find what an excellent 

 form, truly intermediate between S. c. ckinensis and tigrina, they are ; they cannot 

 be united with either of the two nearest forms. The upper wing-coverts have 

 dull black shaft-lines as in tigrina, but narrower ; the edge of the wing is darker 

 than in tigrina, as in ckinensis. The under tail-coverts vary from buffy greyish- 

 white to grey, with more or less distinct buffy tips. I propose to name this well- 

 marked subspecies 



Streptopelia chinensis vacillans, snbsp. nov. 



Type $ ad., Mengtsze, Yunnan, 30. vi. 1910. No. 249, collected by Mr. 

 Alan Owston's Japanese collectors. In the Tring Museum. 

 References are : 



Turtur tigrina Anderson, Western Yunnan Exp. p. 665 (1878). 

 Turtur chinensis Ingram, Nov. Zool. 1912, p. 272. 



Spilopelia tigrina Bangs & Phillips, Bull. Mus. Comp. Zool. Harvard Coll., Cambridge, Mass., 

 lviii. p. 270. 



