The Biology of Senescence 



owe any part of their misfortune to the senile increase in 

 vulnerability, however early in life this is taken to begin. 



Bodenheimer (1938) draws a useful distinction between the 

 'physiological' longevity of a species — that attained under op- 

 timal conditions in a genetically homogeneous population and 

 approaching the longest recorded life-span within the species; 

 and the 'ecological' longevity, which is the mean longevity ob- 

 served empirically under given conditions. The ideal rectangular 

 'physiological' curve postulated by Bodenheimer is a convenient 

 abstraction, at most, since the genetic and environmental con- 

 ditions laid down for it cannot in practice be obtained in any 

 real population. But in some forms the observed life-table in 

 laboratory culture or domestication approximates to the ideal 

 rectangular form, and this approximation is closest of all in 

 some human societies. It is, however, pointless in terms of the 

 actuarial definition of senescence to pursue a 'physiological' as 

 opposed to a 'pathological' senescence in most laboratory 

 animals. If senescence is measured as increased general vulner- 

 ability, Bodenheimer's 'physiological' longevity represents only 

 the approximate region in which the rise in the curve of 

 vulnerability to all assaults of the environment becomes so steep 

 that even major protection against such assaults is insufficient 

 to prolong life very greatly. The pattern can be modified and 

 the apparent physiological longevity increased by removing 

 specific causes of death — e.g. enteritis and ear disease in old 

 rats (Korenchevsky, 1949) but the postponement of death 

 obtainable in this way is itself limited, and argument about 

 'natural' death, apart from pathological processes, in mammals 

 is quite otiose. 



It is manifestly impossible to demonstrate senescence from 

 life-tables unless the mortality in early and adult life is suffi- 

 ciently low, and the number of animals reaching old age is 

 therefore sufficiently high, for an endogenous increase in sus- 

 ceptibility to death-producing factors of random incidence to 

 be evident. Thus wild mice die at a rate which precludes their 

 reaching old age, but mice kept under laboratory conditions 

 have a life-table similar to that of Western European human 

 populations in the year 1900 (Leslie and Ranson, 1940, Fig. 30, 

 p. 109 ; Haldane, 1953) : not even the most cherished labor- 



24 



