148 TRINIDAD AND TOBAGO BULLETIN. [XX.2,3&4. 



Daily Movements. This bird is often seen flying in large 

 numbers in one direction. Careful observation shows that in any 

 one locality these are almost alv/ays in the same direction every 

 morning and in the opposite direction every evening. In fact the 

 birds from a very large area congregate every evening in some 

 special roosting locahty and from there spread each morning to the 

 surrounding country. Two of these roosting places I have located 

 approximately, but have not had an opportunity to see either of them. 



One of them is in the mangrove swamp between Caroni and 

 Port-of-Spain and is located by observations of the birds in 

 Port-of-Spam flying to the west in the morning and to the east in 

 the evening (F. V/. Urich), and by several observations by C. M. 

 Roach of enormous numbers flying to the east between San Juan 

 and Caroni at 5.30 to 6.30 a.m. in August 1920. On August 7 he 

 states that there must have been over 20,000 birds, while on 

 August 14 there were about three times as many as this. These 

 must have been flying from some roosting place near the mouth of 

 the Caroni river. 



The second roosting place is somewhere near the coast near 

 Couva and probably near where the Monkey Point road reaches the 

 sea. Most of the birds from the greater part of the Naparimas 

 return here every night to sleep. From Princes 1 own to the Oropuche 

 Lagoon they can be seen flying northward every evening about 5.30 to 

 6.30 p.m. in great numbers, gradually converging to the coast by 

 Claxton Bay. The reverse flight takes place every morning about 

 5.30 to 6 a.m. and the approximate position of the roost was 

 ascertained on September 17, 1920, while driving from San 

 Fernando to Port-of-Spaln in the early morning. It was then 

 observed that from Claxton Bay (at about 5.45 a.m.) to about half 

 a mile south of the Monkey Point Road all the buds were flying 

 almost due south ; from here to the Monkey Point Road there was 

 a gradually increasing number flying towards the east ; just north 

 of this road all were flying due east, by Couva village many were 

 flying north-east, and a mile further north a few were flying still 

 more towards the north. They were therefore radiating from a point 

 near the coast just north of Monkey Point Road. 



The directions of these early morning flights and the approximate 

 positions of the two roosts are given in the adjoining map which 

 shows the greater part of the east coast of the Island. 



