664 RODENTS AND RABBITS xxv. 5 



extending over many years. Thus figures for the furs collected by 

 trappers for the Hudson Bay Company enable the variations in numbers 

 of the varying hare to be followed back to 1850 (Fig. 436). The cycle 

 is surprisingly regular, with a period of 9-7 years. During the periods 

 of abundance of the hares the mammal and bird predators of these also 

 increase and fall off a little later than the herbivore populations, 

 changes in food taken by the predators producing all sorts of secondary 

 effects on the animals of the area. 



160,000 

 1-1 



j$ 120,000 

 80,000 



3 40,000 



1850 1860 1870 1880 



1890 1900 

 Year 



1910 1920 1930 1940 



1890 1900 

 Year 



1930 1940 



Fig. 436. Fluctuations in numbers of varying hares (above), and lynx (below). 



Estimated from the Hudson Bay Company's fur returns. (From W. J. Hamilton, 



American Mammals, McGraw Hill Book Company. After MacLulich.) 



The existence of these fluctuations is striking evidence that animal 

 populations are not maintained in any stable equilibrium. It is often 

 suggested that the cycles depend on changes in the physical conditions, 

 for instance, on the sunspot cycles, but no close correspondences have 

 been discovered. The sunspot cycle has a period of 1 1 -i years, whereas 

 rodent and lagomorph cycles varying from the 9-7 years for the hares 

 to 5 years for lemmings and 3 for voles have been recorded. Changes 

 in the amount of solar radiation no doubt produce considerable effects 

 on the animals. There is some evidence for a direct effect of diet 

 increasing the fertility. In South America plagues of rats have been 

 observed to coincide with abundance of bamboos. But it seems likely 

 that the cycles of numbers depend on the particular balances set up 

 within the animal communities. These presumably depend on the 

 interactions of the reproductive pressures of the plants, herbivores, 

 carnivores, and parasites, and it is easy to believe that in some condi- 



