THE ROLLER. 35 
themselves, it is necessary to separate them as soon as they are fully grown.” 
A friend of mine attempted to keep some of these birds in an aviary, the 
centre of which consisted of a deep tank, with central fountain, in which fish were 
swimming about; he, however, found it difficult and expensive to obtain sufficient 
fish to supply their needs, consequently they hardly had enough to keep them in 
vigour: in addition to this the cats in the neighbourhood used to clamber over 
the aviary at night making the birds dash frantically about and cut their heads in 
their efforts to force a way through the wirework: thus one by one they got 
drowned, being apparently too weak after a plunge to rise from the water. I do 
not think he kept any of them alive for more than three weeks. 
In the first volume of the Avicultural Magazine, pp. 65-67, is a very interesting 
article by Mr. C. P. Arthur on rearing and keeping Kingfishers; he has probably 
been as successful as anyone with these birds, but he does not recommend them 
as pets for several reasons; one being that these birds never learn to recognize 
the fact that they cannot fly through wire netting, so that they make straight for 
any object beyond it until stopped by the wire against which they flutter helplessly 
breaking their wings and tail feathers. In the second place he says that they are 
liable to fits (for which doubtless unnatural food is to blame); thirdly their cage 
soon becomes offensive; fourthly they are not long-lived; and lastly they have no 
song. . 
Swaysland, who has also been tolerably successful, strongly objects to the 
practice frequently adopted of putting the meat, egg, worms, etc., into a vessel of 
water for the birds to fish out, as he says that the young birds constantly tumble 
into the water, get soaked, and often die from cold. 
FAMILY CORACUDA. 
HE Rollers are birds of large size and brilliant colouring which, on account 
of their somewhat Corvine aspect were formerly associated with the Jays 
(in India they are still supposed to be Jays by the unscientific). Later their 
affinity to the Bee-eaters was pointed out, and Seebohm actually placed them in 
the family Meropide. Huxley associated both groups with the Kingfishers, 
Hoopoes, and Cuckoos, and it has been shown that in the characters afforded by 
their digestive organs and muscles they approach the Nightjars. 
