82 THE RIDDLE OF MIGRATION 



there before and given a sufficiently good memory, 

 no other explanation is necessarily called for. The 

 classical experiments of Watson and Lashley, with 

 noddy and sooty terns nesting on the Tortugas Keyes 

 suggest that in this case homing may have involved 

 something more than memory but even here the 

 evidence is not entirely conclusive. The past his- 

 tory of the birds used was unknown and they may 

 conceivably in previous autumnal wanderings, have 

 covered the ground. Breeding birds were taken by 

 boat to Cape Hatteras (850 miles north) and liberated 

 here and at various points on the way, some of 

 them at sea and out of sight of land. By climbing 

 to the skies on liberation, however, land would soon 

 be observable. At 15,000 feet (the greatest height 

 at which migrating birds — cranes — have ever been 

 recorded) land would be visible at over 150 miles. 

 The sighting of landmarks by these terns can thus 

 not be excluded from the possibilities of the occasion. 

 Yet there is homing of a type from which sight 

 and topographical memory must apparently be 

 excluded. Penguins, for instance, returning to 

 their habitual breeding grounds cannot possibly be 

 making use of landmarks. Their life is spent chiefly 

 among drifting ice-floes. They are flightless and 

 cannot take observations from the air. But, like 

 so many winged migrants, they nevertheless make 



