282 COEACIAS INDICA. 



Head dusky bluish green, brightening above and behind the eye to turquoise-blue ; above the nostril the forehead is 

 greyish yellow, •with a tinge of violet in some ; lower hind neck, interscapular region, and scapulars dull brownish 

 green, separated from the blue of the nape by a vinous collar ; lower back cerulean blue ; upper tail-coverts, base 

 and terminal portion of all but centre rectrices, least wing-coverts, greater part of primaries, and terminal half of 

 secondaries deep violet-blue, with a brilliant cobalt lustre close to the shafts and at the edge of the wing-coverts ; 

 central rectrices dusky green, with a blue wash at base ; a broad band across the remaining rectrices, another 

 across the six outer primaries, primary-coverts, and bases of secondaries pale cerulean blue. 



Lores tawny brown ; beneath the eye and the ear-coverts vinous-brown, with whitish mesial streaks ; throat and chest 

 pale greyish vinous, the feathers with mesial buff lines, and broadly margined on the fore neck and upper part of 

 chest with purple-violet ; beneath, from the chest, with the under wing, pale greenish blue. 



Young. Iris brown, the grey outer portion in the adult reduced to a narrow ring ; this latter increases with age very 



gradually, imparting considerable variation to the eye ; bill blackish brown, pale or reddish at the base beneath ; 



tarsus slightly tinted with olivaceous ; gaj>e yellowish. 

 Head and back duskier than in the adult ; forehead with more of the pale colour ; band across hind neck fawn ; 



lesser wing-coverts (in the nestling) almost concolorous with the back ; chin and throat paler than in the adult, 



the purplish lilac on the latter faint. 



Obs. Ceylonese examples average, I think, smaller than Indian. Two of the latter from Kamptee measure 7'1 inches 

 in the wing ; another 7 - 4 — the former being the maximum limit (according to my experience) of the insular bird. 

 The lilac tints show considerable variation in continental as well as in Ceylonese specimens, the depth of tint 

 depending on age. 



Distribution. — The Roller has a peculiarly local distribution in Ceylon, dwelling in the dry portions of 

 the island, and migrating to the damp district of the west chiefly during the dry season (N.E monsoon). Its 

 head-quarters may be said to be the Jaffna peninsula, the open portions of the northern sea-board, and 

 certain parts of the interior of the Northern and N.W. Provinces. In these districts it is common in many 

 places and absent from others. Neither Mr. Holdsworth nor myself observed it in the Aripu district, but on 

 the adjacent island of Manaar it occurs. To the south of the jungles bordering the coast of the Bay of 

 Kalpentyn it is not uncommon. I have seen it in the Kalpentyn peninsula itself, and about Puttalam and 

 Chilaw it is a well-known bird. It is a resident as far south as Madampe, and likewise in the region between 

 that and Kurunegala ; but below this line it occurs chiefly as a straggler between the months of October and 

 March. In this season it may often be seen about Veangodde and Ambepussa, and I have procured it in the 

 Hewagam Korale, a little to the south of Colombo, in July. I doubt, however, if it resides in that district. 

 I have never seen or heard of it to the south of the Kaluganga, nor did I meet with it in the very likely 

 country between Haputale and Hambantota. It may occur in the Eastern Province, but I have no infor- 

 mation to that effect. In the Trincomalie district it is now and then seen from December to February ; but 

 a little inland, about Ratmalic, it is common enough. Eastward of this point, through the centre of the 

 island, it musters, as above remarked, strongly, confining itself, of course, to open districts, fields surrounded 

 by the village tanks, and dried-up paddy-land. Even here, however, it is local ; for although it is common 

 near Hurulle, I have never seen it about Haborennn, which is separated from the former place by a tract 

 of forest. 



It has been found now and then in the valley of Dumbara, but I do not know that it occurs elsewhere 

 in the Kandyan country. 



On the mainland this species is found throughout nearly all India, from the extreme south to the 

 Himalayas ; it does not extend into Burmah, being there replaced by the closely-allied race C. affinis. The 

 two forms blend into one another in such a gradual manner that it is difficult to say where indica ends and 

 affinis begins. Mr. Hume remarks of Mr. Inglis's specimens from N.E. Cachar, that " they are not very 

 typical, but that they are nearer to typical affinis than to indica." Its range is not by any means so limited 

 towards the north-west ; for in that direction it extends through Persia to Asia Minor, mingling thus with its 

 European ally C. garrula. Mr. Danford observed it in Asia Minor at the base of the Aladagh mountains ; 

 and Messrs. Sclater and Taylor have seen a specimen in the Museum of the American College at Constantinople 

 which was shot on the Asiatic side of the Bosphorus. 



