2 BIRDS OF TUNISIA 



a table giving the differential characters between them. Jlr. Dresser sums 

 up his conclusions at the end of the article as follows : — • 



" Stunucs unicolor may be placed on one side, as it is quite distinct from 

 any other species of Starling. The remaining Starlings may be divided into 

 two groups, viz., those having the wing-coverts green or steely-blue, and 

 those having the wing-coverts purple. The first group contains only 

 S. vulgaris and S. menzhieri [which, as above shown, cannot be separated 

 from S. ruhjaris], and to the latter group belong S. 2}ti'>'pwascens, 

 S. porphyroiiotus, S. poltaratzshji and S. caucasicus. Of these I hold 

 that S. porphyronotus cannot be separated from S. purpurascens, and that 

 both S. poltaratzskyi and S. caucasicus are very close to S. purpurascens, 

 differing therefrom only in being more glossed with green, the former 

 having the back glossed with steely-green, and the rump purple, whereas 

 in S. caucasicus the green gloss extends down to the rump." 



On the whole, so far as I have been able to judge from specimens which 

 I have examined, I am inclined to agi'ee with the above remarks, and to 

 look upon S. menzhieri as a form of S. vulgaris, and upon S. poltaratzskyi 

 and .S. caucasicus as forms of .S'. purpurascens. 



The common Starling is most abundant in Northern and Central 

 Tunisia during the winter months, arriving in vast flocks in October 

 and November, and leaving again for the north in the following 

 spring. Apparently Central Tunisia forms the southern boundary 

 of the migration of this species, at any rate I have no note of its 

 occurrence further south in the Eegency. This Starling, unlike the 

 following species, is strictly a winter migrant in Tunisia, and, so 

 far as I am aware, never breeds in any part of the Regency. In 

 Algeria and Marocco, where it is also plentiful in winter, the same 

 is probably the case. Loche alludes to isolated pairs of the species 

 nesting occasionally in Algeria, but I know of no authenticated case 

 of such nesting in Tunisia. 



Throughout the winter months large numbers of the common 

 Starling are to he met with in many pai'ts of Northern and Central 

 Tunisia, frequenting low-lying water-meadows and marshy plains, 

 and many are netted by the Arabs in the cane-brakes and reed-beds, 

 whither the birds resort of an evening to roost. During this season 

 the market in the town of Tunis is abundantly supplied with 

 Starlings, and being a cheap article of diet they fonu a frequent, if 

 not a very appetising, " plat " in the " menu " of many a Tunisian 

 household. 



Tn many parts of the Continent Starlings are similarly netted and 



