124 SEA-BIRDS 



kinds of small birds. He came to the conclusion that temperature 

 and sunshine play a very important role in timing the breeding seasons 

 of birds, quite apart from the general factors, such as the length of 

 daylight, which zoologists have normally regarded as influencing 

 the breeding season. It appears that the influence of temperature and 

 sunlight, while important, can be regarded more as determining the 

 moment when the trigger is pulled, the cocking of the gun being 

 controlled probably by an internal rhythm, and not, as has been 

 widely supposed, by the hours of daylight. In any case it is doubtful 

 whether climate has ever completely prevented breeding at least in 

 the mature sea-bird, by direct action upon the birds' glands. But 

 climate does often, indirectly, prevent breeding, particularly in the 

 Arctic. It is now commonplace of arctic observation that after late 

 springs and in hard weather some, or even all, of the local bird popu- 

 lation fails to breed. In our opinion these non-breeding years are 

 simply due to the bad weather closing down the food-supply. For 

 instance, in very late springs the ice never breaks up in Baffin's Bay 

 and the sea-birds of its cliffs cannot find food within operational range 

 of their breeding colonies. The same state of affairs undoubtedly 

 happens from time to time in east Greenland and Spitsbergen, where 

 the state of the ice fluctuates a great deal. In King Charles Land, 

 east of Spitsbergen, only nine species of sea- and shore-birds were 

 present in 1889 when the ice conditions were bad, whereas 21 species 

 were reported in 1898, when they were good. In spite of the work 

 of Marshall and others it seems to us that successful breeding in the 

 arctic regions depends primarily on whether the adults can find enough 

 food (first for themselves and later for their young) within easy reach 

 or flying distance of the nesting site. 



To recapitulate previous argument, we find that sea-bird numbers 

 like all animal numbers, depend primarily on food ! So we arrive by 

 analysis at what is, all along, the conclusion which common sense has 

 already indicated! But not all sea-birds are always straining — even 

 in the breeding season — at the limits of their food-supply. We have 

 seen, for instance, that in the last hundred years the numbers of gan- 

 nets in the North Atlantic have been less influenced by the amount 

 of food available than by being preyed upon by man, who does not 

 behave as a natural predator. Under the present protection the North 

 Atlantic gannet is increasing; it may one day return to equilibrium 

 with its normal food supply, a position which may mean that its num- 

 bers will be as high or higher than when man started his ancient 



