B. F. CUMMINGS : BIRD ROOSTS AND ROUTES. 121 



morning is a frosty one they stay on the rookery trees 

 longer than usual. 



At Tapely Park, Instow, Jackdaws collect in pro- 

 digious quantities, numbering many thousands (though 

 it is extremely difficult to judge the number), and roost 

 in the beech trees. A roost of Rooks occupies the same 

 group of trees. The interesting feature connected with 

 these Jackdaws is that the birds, in going to and from 

 their roost, always take exactly the same route. A large 

 flock which, during part of its course, is forced to fly 

 over the town of Bideford, always flies across exactly 

 the same part of the town every evening. It was by 

 watching and following up for several days another big 

 flock (numbering 200 or 300), which fed daily in the 

 fields at Braunton (about three and a half miles from the 

 roost) throughout the whole of last winter, that I finally 

 discovered this large roost. Every morning and every 

 evening this flock as regularly as a Royal Mail performs 

 this journey. They follow very carefully the same line 

 of flight, even to the barest detail, but occasionally they 

 fly very high, and they then appear to follow a more 

 direct course, for it is noteworthy that these birds do 

 not, as a rule, make a bee-line by any means. The reason 

 why they sometimes fly at a great height I cannot imagine. 

 I do not think that it has anything whatever to do with 

 wind or weather. Arrived at the roost, the birds 

 " rocket " down perpendicularly, dropping like plummets 

 through space, and commence to " chock " for an hour 

 or more before darkness falls. Starlings and Wood- 

 Pigeons when dropping in to roost, " rocket " down in 

 this same eccentric way, and many birds behave 

 similarly at times, when they may be said to be " at play." 

 The habit with the roosting birds is, however, a constant 

 one, and takes place every evening. I have found 

 another big Jackdaw roost at Eggesford — in a very 

 wooded district. 



Far more striking evidence as to the use of flight-lines 

 in these " miniature migrations " is to be seen in the 



