CORACIAD^. 215 



The Indian Roller is distributed throughout the whole of India, 

 from Ceylon* and Cape Comorin to the base of the Himalayas ; 

 towards the North-east of our limits it begins to disappear, and 

 is replaced by the next species, and in the extreme North-west 

 its place is taken by C. garrula of Europe and Western Asia. 

 Adams asserts that it is also found in Ladakh and Tibet, and 

 he ignores the existence of C. garrula in the North-west. 



It frequents alike open jungles, groves, avenues, gardens, clumps of 

 trees in the open country, and, except inthick forest, is to be found 

 everywhere, and is sure to be met with about every village. It gene- 

 rally takes its perch on the top, or outermost branch, of some high 

 tree, and, on spying an insect on the ground, which it can do at a 

 very great distance, it flies direct to the spot, seizes it, and returns 

 to its perch to swallow it. A favourite perch of the Roller is a 

 bowrie pole, or some leafless tree whence it can see well all around ; 

 also old buildings, a haystack, or other elevated spot; sometimes a low 

 bush, or a heap of earth, or of stones. When seated it puffs out the 

 feathers of its head and neck. I have, on several occasions, seen 

 one pursue an insect in the air for some distance, and when the 

 winged termites issue from their nest after rain, the Roller, like 

 almost every other bird, catches them on the wing. It flies in 

 general with a slow, but continued, flapping of its wings, not unlike 

 the Crow, though more buoyant ; but it has the habit of occasionally 

 making sudden darts in the air in all directions. Its food is chiefly 

 large insects, grass-hoppers, crickets, mantidse, and even beetles ; 

 occasionally a small field-mouse, or shrew. It is often 

 caught by a contrivance, called the Chou-gaddL This consists of 

 two thin pieces of cane, or bamboo, bent down at riglit angles to 

 eacli other to form a semicircle, and tied in the centre. To the 

 middle of this the bait is tied, usually a mole cricket, sometimes 

 a small field mouse, (Mus lepidusj ; the bait is just allowed tether 

 enough to move about in a small circle. The cane is previously 



* Cabanis gives the Ceylon bird as distinct from the continental one, keeping 

 Indicus for the former and Bengalensis for the latter bird ; but as his specimen of 

 Bengalensis is from Nepal, he has perhaps got a hybrid between this and affinis. 



