328 NOVITATES ZOOLOGICAE XXV. 1918. 



I united " Sturnus sopliiae " with <S'. v. vulgaris and treated all Starlings, with the 

 exception of tmicolor, as subspecies of ;S'. vulgaris. Tlie latter is doubtless, from 

 my point of view, correct, unless S. unicolor be also joined as a subspecies to 

 vulgaris. S. sophiae — of course as S. vulgaris sophiae — is, I think now, the name 

 to be adopted for the East Russian Starhng, if the latter is separable at all. 

 About this below. I described <S'. v. granti from the Azores, united menzbieri 

 and poUaratskyi, kept jaroensis separate, resurrected nohilior, but made the 

 same muddle of purpurascens, though I felt at the time that this was not 

 satisfactory. 



In 1904 ButurUn published (Oni. Jahrh. 1904, pp. 205-213) descriptions 

 of 6 new subspecie.s and 3 new species, and added a " key " in which he recog- 

 nized 22 forms grouped into 11 species ! I am sorry to say that this article, 

 which necessitated a rene\^ed studj' of the whole genus, while acquainting us 

 with a few very interesting new forms, which considerably advanced our know- 

 ledge, contains a number of clear synonyms, and that the grouping in 1 1 species 

 (i.e. unicolor, humii, vulgaris, faroensis, poUaratskyi, zaidamensis, dzungaricus, 

 purpurascens, minor, tauricus, and porphyronotus) is arbitrary and misleading, 

 though ButurUn was, in several instances, obliged to admit several species if he 

 adhered to the dogma that subspecies inhabited separate areas — Ids supposed 

 species, however, are in some cases mere synonyms, because he did not grasp 

 the fact that among Starlings — as in most birds — a certain amount of individual 

 variation exists, and som times, for example in dresseri and poUaratskyi, to an 

 inconvenient degree. Such individual differences can easily be mistaken for 

 subspeciiic characters, but their true nature becomes evident if sufficient series 

 are compared. There is little excuse that Buturhn did not grasp them, because 

 he, apparently, had large series for study. 



In 1905 Tschusi (Orn. Jahrb. p. 141) described Sturnus vulgaris graecus, 

 in 1909 Buturlin and Harms " Sturnus balcanicits," in 1915 Oscar Neumann 

 S. vulgaris oppenheimi. 



Unfortunately I am not able, at present, to compare any of Buturlin's 

 types, nor the type of Neumann's S. v. oppenheimi, but I have good material of 

 topotypical S. v. balcanicus and S. v. graecus, the latter from Reiser. 



The authorities of the British Mvrseum allowed me to compare their 1,100 

 Starlings with our fine series of most forms, and Dr. Wm. Eagle Clarke kindly 

 lent me the Starlings from the Scottish Islands in the Royal Scottish Museum ; 

 a careful study of all this material leads me to the following conclusions. 



Sturnus vulgaris vulgaris L. 

 Vog, pal. Fauna, p, 41, 



It is at present uncertain how far eastwards the typical Starling extends, 

 but we may safely assume that it inhabits Ireland, Scotland, and England, 

 Scandinavia, Central Europe and France to the Pyrenees, Italy, Austria and 

 Hungary, and the western parts of Russia. As I have said before, StarUngs 

 with more or less purple on the head are common even in westernmost Europe, 

 and it seems that some of them are not distinguishable from East-Russian 

 sophiae, but they are not stragglers from the Ural, but residents Uke the 

 green-headed specimen.-;, i.e. mere individual aberrations. 



