122 THE MIGRATION OF BIRDS 



Birds may be carried on ship-board. When the 

 '' Mauretania ' was between 400 and 500 miles 

 out from New York, bound eastward on June 15th, 

 1911, a curlew came on board and remained for 

 three days, leaving when the Irish shores were 

 sighted on the 18th. My informant, an ex- 

 perienced wildfowler, failed to catch the bird, 

 but described it as like our curlew. Probably 

 it was the American Numenius longirostris , but 

 amongst the Irish curlews it would easily remain 

 unrecognised. 



When a seabird appears inland the usual ex- 

 planation given is ' storm-blown," but increasing 

 knowledge shows the frequent fallacy of this idea. 

 The Manx shearwater, for instance, is a regular 

 migrant, and the examination of the dates of the 

 records of so-called " storm-blown *' birds found 

 in inland localities, shows a remarkable regularity; 

 the majority are met with between the end of 

 August and the end of the first week in September. 

 Not only do the birds move south in the early days 

 of September but many, usually at any rate, cross 

 England ; the weaklings fall out and are found. 

 Is it possible that some of these collapses of 

 passing birds are due to more than mere physical 

 fatigue ? Aviators have discovered the existence 

 of ' wind pockets " or " holes in the air," where 

 the resistance of the air appears suddenly to fail ; 



