6 A. Comfort 



definitions which we employ must be of such a kind that they 

 do not beg the questions which comparative studies are de- 

 signed to answer. 



Measurement of senescence 



If we accept the idea that animal senescence may be hetero- 

 geneous, it follows, I think, that the more general the test we 

 use to measure it the better. If senescence is to be treated as 

 I have assumed, its most convenient and general measure is 

 the increase with age in the probability of death. The ad- 

 vantages and reservations of this measure have already been 

 discussed by Medawar (1955&). If senescence is defined other- 

 wise, then other measures will follow from the definition. 

 The actuarial measure is that normally used in man, and it 

 measures the total loss of vigour against all stresses and from 

 all causes. 



Alternatively, we could measure the decline of one particular 

 kind of homeostasis, such as the resistance to temperature or 

 to drugs, or one of the characters contributing to vigour such 

 as fertility or power of regeneration. Most of these indices 

 show changes against age in most animals, but they do so 

 in varying proportions and in contrary senses. Accordingly 

 decline in vigour is often better seen as an increase in the 

 variance, the coefficient of variation, or some other index of 

 scatter for some or all of the characters of vigour, rather than 

 as a parallel trend in all of them. This is true, for example, of 

 the human blood pressure (Burger, 1954), where the individual 

 values diverge throughout life. We might expect a similar 

 increase in the variance if we measured the regeneration-rate 

 of tissue in an animal which displayed a slowing of regenera- 

 tion with age, but which had the property of regenerating 

 previously injured tissues more rapidly than uninjured tissues. 



Increased variability is sometimes the only character of 

 ageing organisms which correlates closely with their rising 

 mortality. This difference between young and old animals is 

 somewhat similar to that between hybrid and inbred animals 

 of identical age. The resemblance may not be coincidental, 



