67 



coast preparatory to starting on their migration to sunnier climes. 

 Often on a tine evening one may see dozens resting on the top of the 

 wire fences along the sea front, telephone wires, etc., while hundreds 

 disport themselves in the air, but by the following day nearly all 

 are gone. Other swarms then gradually collect until they, too, 

 disappear as suddenly. These successions of migrations usually 

 occur from mid-September to mid-October, but one generally sees 

 very few Martins hereabouts after that time. I was therefore not 

 a little surprised, on November 7th, to notice the evident collection 

 of a migratory swarm of these birds on the coast, it was not a large 

 one it is true, and probably included rather less than a score of 

 individuals, nor does the date constitute a record, as migrations have 

 been noted as late as the 16th, bat then records are exceptional, 

 and not often repeated. Probably these belated migrants are late- 

 hatched birds that are not strong enough to start their journey with 

 the earlier groups, and possibly it was such as these that caused 

 our older naturalists to believe that the Martins hybernated in this 

 country. 



