ON ASIATIC STARLING . 38 



especially the crown and throat. It is said to inhabit Eastern Russia, 

 wintering on the Talysh, passes through the Caucasus to Palestine, Cyprus 

 and Egypt. I can only remark that a great many west European Starlings 

 have, as indicated above, the head andthroat purple and I cannot believe 

 that sophiae is a good race. 



Sturnus vulgaris oppenJieimi was described by Neumann and is said to 

 breed in northern Mesopotamia. It is evidently very close to purpurascens 

 from which it is said to differ by having the head glossy green with a 

 purple base, nape steel blue, back blue green and outer webs of secondaries 

 and their coverts with a bronzy gloss ; upperwing coverts, lower back and 

 rump purple, underside more or less purple; the glossy green throat and the 

 colour of the head, nape and rump would seem to distinguish this race too 

 from 7iobilior. 



Sturnus vulgaris dresseri was described by Buturlin. This race appears to 

 be somewhat intermediate between porphyronotus and nobilior. It is said 

 to differ from porphyronotus in having the head and throat more purple. 

 Back, rump and upper tail coverts are more uniform purple but the mantle 

 has " violet to steel blue and even dark green colours, in fact, the back is 

 very variable and occasionally quite deep metallic green " (Hartert). This 

 should distinguish it from nobilior. The wing too is said to be smaller, II 

 specimens measure 125-131 mm. 



The distribution, so far as is known, is to the N. and N. W. of that of 

 porphyronotus. It ranges from Askabad and Merve to Ferghana, and 

 according to Buturlin, east as far as Kara-tau. In winter not rare at 

 Kandahar. In the British Museum there are specimens from Mardan and 

 Sind (Hartert). Possibly this last may be that recorded by Gates from 

 Sind as humii ; this specimen requires re-examination as I know of no 

 other record of humii irom Sind. 



Sturnus vulgaris dzungaricus is described by Buturlin who believes this 

 race nests in Dzungaria. This race seems to be very near nobilior and 

 dresseri. Hartert says that the birds he has seen are too large (wings 

 129-135) for the latter and the underwing coverts too light. As this race 

 has occurred in India (according to Hartert who has examined birds from 

 Meerut, Mardan, Lvicknow, and llawal Pindi), I append the distinctions 

 he gives though I must confess that on paper they are very difficult to 

 visualize. 



" The head is green with a strong purplish gloss, or, as Buturlin says, 

 * bronze purple,' the back is purple, interscapular region, however with 

 more or less steely -blue green, edges of the wings bronze with more or 

 less purple gloss." 



One of the above birds (the Rawal Pindi one, Biddulph coll.) was labelled 

 by Sharpe as purpurascens (Fauna i, 524). 



Stur7ius vulgaris zaidamensis was described by Buturlin from 2 specimens 

 from Zaidam and Ta-tschu, N. of Nar.san Mts. It appears to resemble 

 poltaratsl-yi but has a green head, neck and throat. 



Two specimens are not sufficient to separate new races of Starlings on. 



General Remarks. — From the above descriptions I think one should be able 

 to name the majority of specimens ; however in first winter dress when the 

 plumage is very spotted and the glossy reflections masked, especially in 

 females (which are always duller than males), considerable difficulty may 

 be experienced and a few will be impossible to name. Very rough skins 

 are difficult and sometimes impossible to differentiate as may be readily 

 imagined when one realizes that the colour sheen depends on the reflection 

 of lisht from the surface of the feathers. 



8-a 



