CAUSES OF ANIMAL CYCLES 187 



eluded that human fertility has become a problem in itself largely 

 divorced from the problem of mortality." This indicates that there 

 is a certain amount of regulation enforced upon the human popula- 

 tion, giving it some of the properties of a climax species. The assump- 

 tion that it will be necessary to use every square foot of land to grow 

 food for coming millions overlooks the possibility of a fundamental 

 change in human fertility from crowding. The well-known cyclic 

 changes in death rate are probably related to weather, nutrition, and 

 disease, but the increase in the world's population has been connected 

 with the discovery of new lands, the industrial and mechanical revo- 

 lutions. The last two developments mentioned have made crowding 

 possible, but at the same time, have brought a measure of control of 

 housing, nutrition, and disease. 



Physiological Changes in Vigor. Changes in reproductive vigor, 

 especially, were experimentally demonstrated in the chinchbug (Shel- 

 ford, 1931, a). Individuals collected out-of-doors each spring and bred 

 in cages during ten consecutive summers showed maximal reproductive 

 vigor in 1919 and again in 1926, 1921 being a year of extreme weak- 

 ness. No explanation for the phenomenon was evident. Green (1932) 

 further suggests a change of tularemia from a virulent to an immuniz- 

 ing stage in the rabbit population as the chief cause of variation in 

 game populations in INIinnesota. Changes already noted in the num- 

 ber of eggs shed by mammals perhaps belong here. Graham (1929, b) 

 correlates mouse abundance with the condition of the larch sawfly. 



Cycles in salmon have long been noted. Initial shortage of repro- 

 ducing population may be suggested to explain the small ''runs" of 

 sockeye salmon in the three years between the "big runs." The "big 

 run" population dies of old age, but leaves either a more numerous 

 or physiologically superior progeny which matures another equally 

 large population at the end of three years (see Babcock, 1908-1914). 

 Because of enormous fecundity of fishes, this is an hypothesis requir- 

 ing more investigation, but the case of the mammals is different. 

 Let us assume that mice die of old age when about three years old. 

 A favorable year may lead to a saturation of available space. Sur- 

 vival of young will then be limited to the space left by accidental death 

 of "favorable-year" individuals and migration out of the area. In the 

 third year, at a time when reproductive capacity is reduced by senility, 

 more than half the mouse population dies of old age, causing a sharp 

 decline. Recovery could start only the following year. This hypothe- 

 sis cannot be tested until some method of ascertaining the age of 

 mammals is determined. The most promising field here is studies of 

 teeth (Schour, 1936; Schour and Poncher, 1937; and Steadman, 1935) 



