296 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXVIII. 
(2) caucasicus is of course a very recognizable race with the red purple wing- 
coverts and green head, throat, mantle, and undertail coverts, when quite 
typical ; however as with other races it is a little variable and T have examined 
two from the type locality (breeding) which had the throat purplish and one 
with purplish violet under-tail coverts. The crown however is invariably dark 
green. 
(3) These three skins are quite inseparable from typical nobilior from Kan- 
dahar, differing from caucasicus in having the head, throat, and undertail 
coverts purple. 
(4) These two birds are very puzzling and agree with no race of which I have 
been able to see descriptions of specimens of. There are two other skins, 
which evidently belong to the same race, collected by Mr. Hotson near Shiraz. 
These four specimens differ from caucasicus in having a bright, vivid green gloss 
on the wing-coverts, mixed with violet-blue, instead of a purple red gloss ; 
also the upper parts from mantle to rump have a varying amount of purple 
sheen mixed with the green, so much so that an adult male from near Shiraz 
has the whole of the upper part bright purple and no green gloss at all. All 
four skins are winter specimens. As the breeding quarters of this bird are not 
known and some races have been described by the Russians, the descriptions of 
which I have not been able to see, I shall not compete with them in making 
another new race! Dr. Hartert who kindly examined these skins for me says 
they are “only caucasicus in their usual garb,” though he admits one has far 
more purple on the back than usual. With this I must, I fear, disagree. I 
have examined about a score of typical cawcasicus and all have the plum-red 
wing-coverts, utterly different to those in these birds ; had there heen one bird 
alone which differed from typical caucasicyg, one might have put it down as an 
aberration, but with four skins all shewing the same differentiating characters 
one cannot help supposing that they belong to some other race. 
(5) One obtained by Cheesman from a flock at Bagdad on November 21st, 
1920, clearly belongs to this race. Perhaps not uncommon. 
This does not exhaust the races of Starling which have been recorded from 
Mesopotamia! Meinerzhagen (Ibis 1914, p. 389) saw purpurascens in abund- 
ance! It may of course occur, but none of our thirty-five skins are referable 
to this race, and one cannot tell it in the field ; such “‘ records ”’ are better left 
unpublished. Neumann (J. F. O., 1915, p. 121) described a new race from 
Northern Mesopotamia which he cails oppenheimi, and remarks that it does not 
fit in with any race according to Buturlin’s key! He says it breeds in Meso- 
potamia as Dr. Pietschmann obtained on May 23rd at Mosul a grey, scarcely 
fledged, young one. Now Neumann’s type came from Tel-Halaf (Ras-el-Ain) 
(between Mosul and the Syrian boundary) in January and he has seen a 
similar bird from Mosul, also in January (recorded by Sassi as nobilior), but 
how he arrives at the conclusion that his new race breeds there is beyond my 
comprehension, especially as the only breeding birds he has seen from this area, 
and which he describes, do not at all fit in with his diagnosis of oppenheimi ! 
What exactly oppenheimi is, it is impossible to say without seeing the type and, 
until a series of breeding birds can be seen from Mosul to ascertain whether a 
distinct race does breed there, it is best, I think, not to recognize Neumann’s 
race. From his description it is very close to purpurascens which breeds in Asia 
Minor and Armenia. Hartert states (Nov. Zool. 25, p. 332) that he has seen a 
Starling from south Mesopotamia which agreed with Neuman’s description. 
7. Rosy Pastor. Pastor roseus. 
Pastor roseus (L.) (Syst. Nat. Ed. X, p. 170—Lappland), 
The migrations of the Rosy Pastor would seem to avoid for the most part the 
Mesopotamian plain. Pitman saw several at Feluja on May 2nd and obtained 
one, and on May 7th saw several more which departed in N. W. direction. On 
