334 XOVITATES ZOOLOGICAE XXV. 1918. 



variable and cannot serve to distinguish between two races; moreover when 

 nammg menzhkri Sharps probably compared chiefly, if not alone, the fine old 

 nobilior, which formed the bulk of his supposed poUaratshji. 



Fortunately I was able to confront the type of Slumus j)ollaratsIcyi, and 

 that of S. menzbieri, the latter formerly in the Seebohm Collection. There is now 

 a specimen from Gurgaon, India, in the British Museum with the foUo^^■ing 

 remark on the label : " This I beheve to be the type of S. menzbieri.'''' Mr. Chubb 

 and others who are most competent agree that this is in Sharpe's handwriting. 

 I beheve they are right, though the writing is to my mind not quite typically 

 Sharpean, and why does he, the author, " beheve " that it was the type ? But 

 even if he had most emphatically written that the Gurgaon bird was his type, 

 this cannot be accepted, because nothing of it was said in the original 

 description, and the remark was evidently not written until after 1903, when I 

 had searched in vain for the type of Sturnus menzbieri in the British Museum. 



Professor Suschkin, " Die Vogelfauna des Minussinsk-Gebietes, des west- 

 lichen Teils des Sajan Gebirges und des Urjanchen-Landes," in Bull. Soc. Imp. 

 Nat. Moscow, " 1912," " 1913," p. 259, separates "Sturnus vulgaris menzbiiri,'' 

 nesting in the Minussinsk area and the plains of Urjanchen-Land, and " Sturnus 

 vulgaris poltorafzkii " (here is another spelling !) as nesting in the southern Altai 

 and on the Kara Irtish, but he states no reason for his theory ; probably he 

 followed Sharpe, having seen a few specimens confirming this assumption. 



Sturnus vulgaris pollaratskyi has somewhat paler under wing-coverts than 

 S. V. vulgaris, and with wider pale buff edges. The upper wing-coverts vary 

 from green to purple, but never become so deep and fine reddish purple as in 

 adult nobilior in the summer. 



The exact distribution of this form is not yet certain. I believe that it 

 breeds from Krasnoyarsk to the Marka-Kul and Saissan-Nor and eastwards to 

 Lake Baikal. Specimens obtained at Kuldja were probably on migration, as 

 S. V. porphyronotus, which occurs there as well, appears to be the nesting form 

 of Kuklja. Sharpe mentions a specimen in the British Museum shot at Kuldja 

 in August, but the date may not be correct, moreover, being most hkely a 

 Russian date, would be from the end of the month ; the skin was purchased 

 from Schliiter and has no original label, and any date may be doubtful if there 

 is no original label. This Starhng winters in great numbers in India and ranges 

 westwards to Persia. 



Sturnus vulgaris porphyronotus Sharpe. 



Sturnm pnrph/ronotus Sliarpe, Ibis 1888. p. 43S (Turkestan, typical locality Yarkand) ; Cat. B, 



Brit. Mils. xiii. p. 38. 

 Sturnus vulgaris porphyronotus Hartert, Vog. pal. Fauna, p. 40. 

 Sturnus porphyronotus loudoni Buturlin, Orn. Jahrh. 1904. p. 211 (" Ostlicher Teil des russisclien 



TurUestans, Kuldscha, Ferghana und verfliegt sich bis Samarkand— nicht typische Vogel — und 



Dscharkend.") 



Easily recognized by the entirely purple back, rump, and upper tail-coverts, 

 but back sometimes with a more pure dark blue tinge. Head and throat bronze 

 greenish, rarely with a purplish tinge, but ear-coverts alwaj's with more or less 

 distinct purple sheen. Smaller than pitrpurascens. Wings of 38 specimens 

 125-133, once 134, once 135 mm. The young are rather light sandy or rufescent 

 brown. 



