8 P. B. Medawar 



The curve in Fig. 1 represents the faUing away of the num- 

 ber of survivors (/^) in each year x of age. Its slope is the death 

 rate, i.e. the rate of dechne of the number of survivors; where 

 it is steep, the death rate is high; were it horizontal, the death 

 rate would be zero. The slope is downward and therefore has 

 a negative sign; to make the death rate positive, it may be 

 written 



1 1 ^^h 



death rate= — -r^ 



ax 



The death rate itself is almost useless as a measure of the 

 likelihood of dying, for its numerical value clearly depends 

 upon the number of individuals still left to die. (The death 

 rate at 100 is lower than at any previous age, because so few 

 are left in the running.) What is required is the age-specific 

 death rate or force of mortality, the death rate at any age 

 divided by the number still exposed to hazard, i.e. 



force of mortality = — - — ^, 

 •^ l^ dx 



the negative sign being used, as before, to give the force of 

 mortality a positive value. In a population of which the mem- 

 bers do not deteriorate with increasing age — e.g. a population 

 of plates or test tubes — the likelihood of being broken is no 

 greater at any one age than at any other; at the end of each 

 age-interval the number of survivors falls to a constant 

 fraction of its value at the beginning of the interval; the force 

 of mortality is therefore constant. The curve of the force of 

 mortality in human beings has the shape illustrated by Fig. 2. 

 It is high at birth, falls steeply to a very low value between 

 the ages of eight and fourteen, and then rises without singu- 

 larity or inflexion for the remainder of life. 



The question now is: can the force of mortality be used as a 

 measure of the degree of senescence over that period of life 

 in which the force of mortality is increasing? (There is no 

 question of its being used for such a purpose during that- 

 earlier epoch of life in which its value is falling.) 



