81 



5 



for this separation. As the various species have been so continually confused, and especially as 

 one finds so frequently one of the other species doing duty for Turdus varius in collections, I 

 give the following table showing the differences by which they may be at once distinguished. 

 This table is drawn up according to a plan arranged by Mr. Seebohm (who has worked out these 

 Thrushes with me), which appears to me to be extremely simple and practical. 



(Fourteen tail-feathers 1 



(Twelve tail-feathers 2 



| Second primary much longer than the fifth 3 



' (Second primary much shorter than the fifth 4 



(Basal portion of the inner web of quills white 5 



'(Basal portion of the inner web of quills buff . . 6 



3. Turdus varius, Pall. (Zoogr. R.-A. i. p. 449, 1811), range as above. 



4. Turdus horsfieldi, Bp. (Rev. de Zool. 1857, p. 205). Java. 



| Underparts entirely buff lunulated with black 7 



' (Underparts white lunulated with black, and washed with ochraceous on the breast . 8 



(Upper parts obve lunulated with ochraceous and black 9 



' (Upper parts olive lumdated with black 10 



7. Turdus imbricatus (Layard), Ann. & Mag. Nat. Hist. 1854, p. 212. Ceylon. 



8. Turdus nilgiriensis, Blyth (J. As. Soc. Beng. xvi. p. 136, 1847). Neilgherries. 



9. Turdus dauma, Lath. (Inch Orn. i. p. 362, 1790). India. 



10. Turdus lunulatus, Lath. (Ind. Orn. Suppl. p. 42, 1801). Australia. 



Besides these six species there is another, which has been described by Mr. Gould from 

 North Australia as distinct, Turdus iodurus (Gould), Ann. & Mag. Nat. Hist. 1872, vol. ix. 

 p. 401, and which may possibly be a valid species; but having only had an opportunity of 

 examining one specimen, I cannot speak with certainty. It very closely resembles Turdus nilgi- 

 riensis, but has the second primary about -02 inch longer than the sixth, whereas in Turdus 

 nilgiriensis it is only '01 inch longer, and the upper parts of Turdus iodurus are more ochraceous 

 in general tinge of colour than is the case in Turdus nilgiriensis. 



In a series of specimens we find also that it is by no means easy to draw a hard and fast line 

 between Turdus dauma and Turdus lunulatus. In the latter species the upper parts are usually 

 much darker and more olivaceous, the feathers seldom having the subterminal ochraceous bars so 

 conspicuous in Turdus dauma, or else they are dull rufescent instead of ochraceous. The central 

 rectrices in Turdus dauma are also much lighter, compared with the lateral ones, than is the case 

 with Turdus lunulatus. In immature examples of this latter species, however, the upper parts 

 are paler, and there are subterminal bands on the feathers of the crown ; so that they approach 

 dark-coloured specimens of Turdus dauma in appearance. Turdus horsfieldi, described by Bona- 

 parte in May 1857, was simultaneously described by Sundevall (J. f. Orn. May 1857, p. 161) 

 under the name of Oreocincla malayana ; but as I well know from experience that the ' Journal 

 fur Ornithologie ' is rarely issued until" two or three months at least after the date on the title- 

 page, I think that there is every reason to suppose that Bonaparte's name was first published ; 

 and I have adopted it accordingly. 



In the above list I have only included those which have the upper parts varied, excluding 



4 m 2 



