64 THE RIDDLE OF MIGRATION 



tain their health in years of reduced ultra-violet 

 radiation, (e.g., persistently cloudy summers) they 

 must (a) change their diet, (b) move further south 

 or (c) suffer the consequences. Adoption of the 

 first alternative is extremely unlikely. Birds ex- 

 cept when rearing young show remarkable fixity 

 in their diet, though it must be admitted that even 

 a dog may instinctively eat grass under certain 

 conditions. An overwhelming majority of northern 

 birds go south for the winter and so automatically 

 fulfill the second alternative. The last descends 

 like a guillotine on a small number every decade. 



Having seen something of conditions obtaining 

 in the northern hemisphere as it exists today, we 

 may now turn to a consideration of certain aspects 

 of the past. 



Undoubted bird-remains go back in the record of 

 the rocks to the epoch known to palaeontologists as 

 the Jurassic, a matter of some seventy million years. 

 Two examples of the first-known birds are in exist- 

 ence, the one (Archaeopteryx) preserved in London, 

 the other (Archaeornis) in Berlin. Both were ob- 

 tained from slate quarries in Bavaria. They differ 

 so radically from modern birds, that were it not 

 for the feathers, of some of which very perfect im- 

 pressions remain in the slate, one would probably not 

 suspect them of being birds at all. The tail is long 



