HEREDITY, VARIATION AND DEATH 37 



for reproduction. For though, if the species is to succeed in 

 the struggle with others, slight changes must constantly be 

 originating in them, yet the animals which grow from them 

 must not deviate widely from type. One of the higher animals 

 is, therefore, a contrivance resulting from Natural Selection for 

 utilising the two tendencies of cells to maintain their characters 

 when they divide, but to vary slightly. To put it more cor- 

 rectly, those individuals, in different parts of which these two 

 tendencies were highly developed, have been selected for sur- 

 vival. But even in the somatic cells, in spite of their returning, 

 when once specialised, to the system of equal fission, as far as 

 equality is possible, inevitable specialisation must lead to the 

 same results that we have seen in one-celled creatures : through 

 repeated fissions they will, at last, no longer have that combina- 

 tion of elements on which life depends. So that both in the 

 Protozoa and the Metazoa we may regard fission and over- 

 specialisation as the cause of death : amphimixis and generalisa- 

 tion as the cause of rejuvenescence. But the multicellular are 

 in a very different position from the unicellular. For the Proto- 

 zoon there is a vitalising process possible, which may restore its 

 youth. For the Metazoon, as a whole, there is no such magic 

 possibility, but only for special cells set apart for reproduction ; 

 for what we think of as the individual, death is inevitable. 



If over-specialisation is the cause of death, and if conjugation Bacteria 

 exists primarily as an antidote to the destructive tendency, how 

 is it that for bacteria in which fission proceeds, as far as is 

 known, endlessly, uninterrupted by conjugation, there is no 

 natural death ? Though I have no wish to shirk answering 

 this question, I must point out that it is quite possible con- 

 jugation does go on among bacteria, though no one has yet 

 observed it. 



But, assuming that it does not, we may certainly hold that 

 for beings that are very little specialised, it is not a necessity. 

 Cell-division may continue for many millions of generations : 

 their living protoplasm may be capable of an eternity of fissions 

 without losing the necessary structural organisation. This 

 view supports the theory which regards death as the result 



