The Biology of Senescence 



an individual. Our estimate of it is determined statistically, 

 upon a population. The demonstration of such an increase in 

 vulnerability is a necessary condition for demonstrating sen- 

 escence: it is, obviously, only a sufficient condition if selective 

 mortality from age-distributed external causes is ruled out. 

 Real populations are subject to mortality both from random 

 and from age-distributed causes — the variation of exposure rate 

 throughout life is familiar in man; grown men are subject to 

 risks which do not affect children, and so on. Differences in 

 'risk' throughout life have been studied in some other animals, 

 such as the locusts whose causes of death were analysed by 

 Bodenheimer (1938) or the gall-fly Urophora (Varley, 1947). 

 Pearson (1895), in his mathematical analysis of the curve of 

 human survivorship into five components, attempted to limit 

 the meaning of 'senile mortality' to one such component, reach- 

 ing its maximum incidence between 70 and 75 years of age. 

 This would be an ideal solution if it were practicable, but 

 Pearson's analysis is artificial in the extreme, and his 'five 

 separate Deaths' directing their fire at different age groups are 

 not biologically identifiable. In general, however, a progres- 

 sively increasing force of mortality and decreasing expectation of 

 life in a population, if significant variation in exposure rate can 

 be excluded, is evidence of the senescence of its individual 

 members. The preliminary test for senescence in an animal 

 species depends, therefore, on the life-table of an adequate 

 population sample, studied with suitable precautions against 

 selective causes of death. 



The expected differences in behaviour, and form of life-table, 

 between populations which age and which do not age are 

 shown in Figs. 5a,b. In a population not subject to senescence 

 and exposed only to random overall mortality, the decline of 

 numbers is logarithmic, and animals die, ex hypothesis from 

 causes which would have killed them at any age. In a popula- 

 tion exposed only to death from reduced resistance, due to 

 senescence, the curve approaches a rectangular form: after a 

 certain age, animals die from causes which would not have 

 killed them in youth. In one case the force of mortality is con- 

 stant, in the second it rises steadily with age. Thus in rats the 

 force of mortality rises after the ninth month of life in a geo- 



18 



