BIRD LIFE OF BARTICA 97 



four due north from the distant forest, and three — probably 

 a single family — flying west from Keow Island. Two 

 months ago, there was a colony of nine inhabited nests in a 

 single tree on this island, so this trio of birds must have been 

 strongly affected by some reason or habit, to desert their 

 brethren wherever they may have roosted, and fly to so gi*eat 

 a distance every evening. 



An advance in roosting ceremonial was presented by the 

 smooth-billed anis. Shortly before sunset w^hen these birds 

 were scattered about the clearing, they ceased feeding and 

 in the small flocks or family groups in which they are always 

 found, they collected at the tops of the low shrubs. Here 

 they climbed about in their aimless way, fluttering awkward- 

 ly, whaloophig to one another, until dusk had begun to close 

 down. One by one they then began their loose- jointed, 

 steady, beating flight, and if another lot of anis was seen 

 below, the first group would often stop and alight. Then 

 there would be more confusion, more chatter, more aimless 

 clambering about the branches, around and even over each 

 other. At last the roosting flight would be taken up again, 

 this time direct, and the flock of black cuckoos would tum- 

 ble headlong into the bamboos, to begin another period of 

 fuming and perpetual readjustment, before darkness put an 

 end to the day's activity of these weird, inexplicable 

 creatures. 



To parrakeets, going to roost was a rite, not to be 

 performed singly as with the caciques, nor lightly and with 

 lack of dignity ani-fashion. Toward late afternoon the 

 small companies of these birds, which since morning had 

 been alternately feeding and screeching high up in the tree- 

 tops of the jungle, ceased from their two chief activities and 

 rising as if at the word of command, whirred swiftly toward 

 some unusually high tree. I found three of these junctions 

 or assembling places, and at first, thought I had discovered 

 the real roost of these birds. Two of the trees were gigantic 

 moras, whose topmost leaves must have been little short of 



