360 FALCONID^. 



to Bengal, especially at the base of the Himalayas, and stragglers 

 have been met with throughout the Peninsula, there being one 

 from Mysore in the Hume collection. 



(2) The race found in Assam, Cachar, and Burma is by Hume, 

 rightly, I think, identified with the Hainan S. rutherfordi. The 

 breast is generally barred, though less distinctly than in typical 

 S. cheela, and the barring becomes less distinct in many Southern 

 specimens. The chin and throat are dark ashy grey. As a rule, 

 there is a second pale tail -bar, less distinct than the posterior one, 

 just beyond the coverts; wings 17 to 18 inches long in Assamese and 

 North-Burmese birds, in Tenasserim specimens the whole length 

 is 23 to 28, tail 9-5 to 11, wing 15-25 to 18, tarsus 3*5 to 3-02. 



(3) The Andaman race S. davisoni, specimens of which have 

 also been obtained in the Nicobars. This is simply a miniature of 

 S. rutherfordi. Length 22 to 24, tail 10, wing 15, tarsus 3-7, bill 

 from gape 1*7. Some Tenasserim specimens seem undistinguishable. 



Fiu'ther south, in the Malay Peninsula, Sumatra, Java, and 

 Borneo, is another small form, S. i^allidus, Walden (Ibis, 1872, 

 p. 363 ; Cat. B. M. i, p. 290, pi. ix), without any bars on the 

 breast ; wing about 14. 



(4) The common form throughout the Indian Peninsula differs 

 from typical 8. clieela in wanting the barring on the breast, or 

 only exhibiting the merest trace of it, in rarely having any black 

 on the chin, and in often having two bands on the tail, the upper 

 band being more or less indistinct. To the northward this race 

 is but little smaller than S. cheela, and has a wing of about 18, 

 but to the southward and in Ceylon the wing varies from 14*5 

 to 17. This form is S. melanotis; the Ceylon race has been 

 distinguished as S. spilof/aster. If we had only this race and 

 typical >S^. cheela to consider they might well be separated, but 

 some small South-Indian and Ceylonese birds are, as Gurney has 

 shown, not distinguishable from specimens from the Malay 

 Peninsula and Sumatra, and these pass gradually into the large 

 northern /S', cheela through S. rutherfordi. This case is therefore 

 similar to that of Caprimulgus macrurus (p. 189). Probably 

 Spilornis cheela offers the most remarkable instance in Indian 

 birds of the diminution of size to the southward. 



Habits, ^c. This Eagle is usually found on trees near water, 

 especially the fine trees along irrigation-channels and canals in 

 Upper India, and along stream-beds in the lower Himalayas and 

 in the Central Provinces and Southern India. It may also be 

 seen soaring and may be at once recognized by the strongly 

 marked bars on its wings and tail, and by its loud plaintive cry, 

 which it frequently utters on the wing. It feeds on snakes, 

 lizards, and frogs, occasionally on small mammals, on birds, and 

 on insects. It breeds on trees in March, April, and May, making 

 a nest of sticks about 2 feet in diameter, lined with a pad of green 

 leaves. It lays usually one egg, sometimes two, more or less 

 densely streaked and spotted with brownish red and purple, and 

 measuring about 2"78 by 2*2. 



