RELATlVli: VIADILITY IN MAMMALS AXD BTRDS. 547 



Obviously where there was a heavy incidence of death on the 

 young, a longer life would be required to secure replacement of 

 the parent by successfully i-eared young. 



Weismann {t. c. p. 11) brought together a very i-e mar kable and 

 cogent series of cases showing that a constant correlation existed 

 between the reproductive powers and the duration of life of the 

 individuals of a species, and his theory has an extremely im- 

 portant place in the histor}^ of the subject. It is to be noted, 

 however, that although he appreciated the probability that a veiy 

 large number of deaths were due to external causes (enemies, 

 diseases, accidents, untoward conditions) he did not exjjlicitly 

 work out the important relation to his theory of Lankester's 

 distinction between average specific longevity and potential 

 longevity. It is clear that if the duration of life be fixed by some 

 internal cause such as the limitation of cell-reprodviction, it must 

 be the potential longevity and not the average specific longevity 

 that is affected ; but if the majority of animals perish from 

 accidents from without, it is difficult to see how their potential 

 longevity could be influenced by selection. 



Metchnikofi''s most interesting work has still further increased 

 the ditiiculty in the way of accepting "VVeismann's theory. 

 Metchnikoff investigated the causes of death in a very large 

 number of cases, including those of insects which die very soon 

 after having laid their eggs, and came to the conclusion that it is 

 extremely doubtful if natural death occurs except in the rarest 

 possible cases. Most animals peiish long before they have reached 

 an age at all approaching what may be regarded as their potential 

 longevity. In the cases that survive the accidents of youth or early 

 maturity, another series of accidents take eflfect. The changes of 

 senility are induced from without ; the various degenerations 

 are brought into existence by the absorption into the system of 

 various kinds of poisons, and these wliether they are due to the 

 exudations of the microbes of intestinal putrefaction, or to those 

 of special diseases, are entirely external to the constitution. 

 Senile animals perish because they can no longer resist common 

 diseases or common accidents. It would be only in a world 

 where the most perfect hygiene reigned, and from which the 

 bacteria and microbes of diseases had been driven, that the 

 majority of the members of a species would attain the potential 

 age, and that death would come on them from purely constitu- 

 tional causes. And so the conception of the duration of animal 

 life being an adaptation to breeding habits, produced by the 

 operation of selection, disappears. 



I cannot doubt but that the average specific longevity is 

 the dominating factor in animal life. At every stage of their 

 existence animals are assaulted from without by enemies, diseases, 

 and their whole environment. To say that is merely to restate 

 the existence of the struggle for existence, the fundamental 

 proposition of the theory of natural selection. If favourable 



