THE PROCESS OF MIGRATION 125 



and Indian Ocean as far south as the Mozambique 

 Channel. Though perhaps outside the scope of a book 

 on local ornithology, an example or two of the experiences 

 on which these conclusions are based, may be instanced. 



With a single exception (and that on the short run to 

 Norway), we noticed land-birds at sea on each of 

 fifty-six voyages — even in mid-Atlantic, snow-buntings 

 boarded our ship. But the numbers observed, save in 

 two cases, were quite insignificant, varying from an odd 

 bird or two up to a few dozens or scores at a time, during 

 an entire voyage. Note also, that in the majority of cases 

 these travellers sought refuge on the ship, and that many 

 of them died there, of cold, hunger, and exhaustion. 



But where were the "acres and square miles " of birds 

 that were not exhausted? These are not seen. The 

 inference is that at sea practically nothing is visible of the 

 real process of migration. All one does see, is merely 

 casual external evidence ■ of its progress — its waifs and 

 strays, its lost wanderers, that, in last resource, seek 

 fatuous safety on passing vessels as drowning men are 

 said to catch at straws. The few that are seen are 

 a bagatelle, a wholly negligible quantity as compared 

 with the unseen armies above. Those seen, moreover, 

 mostly perish. 



Even those birds that have so far succeeded in their 

 over-sea journey as to have reached the outermost light- 

 ships — should they be so far deceived by them as to drop 

 their flight earthwards — will still assuredly perish, should 

 adverse conditions be then encountered. They cannot 

 cover even that short distance unaided. 



The following is an illustration. On April ioth (1906), 

 we were traversing the Eastern Mediterranean, homeward- 

 bound. While of! Crete (which was dimly visible, 15 



