132 ANIMAL ECOLOGY 



always easy to discover. . . . We have written this paper to 

 draw attention to the variations which take place in a limited 

 area. It is a line of investigation which we think would well 

 repay greater study, and which, if pursued in other areas 

 showing different conditions, might yield sufficient data to 

 make it possible to draw definite conclusions." 



7. Irregular factors in the environment may sometimes 

 act at long intervals but still have a tremendous effect on the 

 numbers of animals. Wood-Jones ^^^^ says that the Cocos- 

 Keeling Islands were visited by very severe cyclones in the years 

 1862, 1876, 1893, and 1902. These cyclones wrecked the settle- 

 ment on the islands and caused much destruction to plants 

 and animals. Similar destructive effects are produced by 

 occasional drought years, which completely dry up ponds of 

 less than a certain depth, causing partial or complete extinction 

 of many species of aquatic animals in the ponds. Each drought 

 is followed by a period of recovery, and it may take several 

 years to reach normal numbers once more. 



8. The irregular changes in the environment may be, on the 

 other hand, favourable to the increase of animals, and so cause 

 unusually large instead of unusually small numbers. For 

 instance, the shipworm {Teredo navalis) was able to increase 

 and spread in Holland in the years 1730-32, 1770, 1827, and 

 1858-59, owing to dry summers rendering the fresh- water 

 regions more saline than usual .^^ It is sometimes the plant 

 environment which changes suddenly. In tropical regions, 

 e.g. India, there are certain species of bamboos which flower 

 only at long intervals, and often simultaneously over large 

 areas. This flowering gives rise to enormous masses of seed, 

 and the unusual food-supply allows various species of rodents 

 to increase abnormally, and sometimes to reach the dimen- 

 sions of a serious plague. ^^ 



9. Periodic fluctuations in numbers may be studied most 

 easily in mammals, since we possess rather accurate and 

 extensive data about certain species whose fluctuations are 

 extraordinarily regular in their rhythm. The best-known 

 cases are among rodents, and of these the most striking are the 

 Norwegian lemming (Lemmus lentmus) and the Canadian 



