96 PROBLEMS OF EVOLUTION 



decide is whether there is an inherent tendency to diminution. 

 Highly as I estimate Professor Llyod Morgan's services to 

 science, I cannot think that he has helped to clear up the 

 question of pammixis. 



Most biologists are unwilling to admit that the predominant 

 tendency is towards reversion. They prefer to assume that 

 apart from Natural Selection retrogression is as much out of the 

 question as the evolution of more complex forms. The favour- 

 able variations, they say, balance the unfavourable. Assuming 

 this, it follows on mathematical principles that a cessation of 

 Natural Selection will lead only to a decline from the survival to 

 the birth mean. 



In the chapter on Heredity and Variation I have argued that 

 variability must inevitably result from the process of multiplica- 

 tion by fission and that the strength of heredity is due to Natural 

 Selection. I picture to myself a species as struggling to ascend 

 an inclined plane : the majority of the component units slipping 

 back, some few only advancing upward, so that only by the 

 aid of stringent selection can the species keep up to the level 

 attained by the successful in the last generation, only by the aid 

 of increased stringency can it climb higher. 

 The main- It may be difficult to believe that an enormous amount of 

 elimination can go on without producing further evolution. Yet 

 this would seem to be the fact when there is no change in the con- 

 ditions. When the environment changes for a species, any 

 individuals that happen to have adaptive variations are singled 

 out for survival. A new standard is set up and those that survive 

 are already up to it. More than a bare superiority to conditions 

 Natural Selection is powerless to bring about, for the species as 

 a body, though, through variations, occasional individuals may 

 rise somewhat above this level. But among the offspring of 

 those selected when the new standard is set up, will be many 

 that fall much below it. The new characteristics will not be 

 constant at first, and Natural Selection will be busy producing 

 fixity, or checking the tendency towards reversion, the tendency 

 to fail to recapitulate the final stages of the phylogeny, to lose 

 suddenly structures that have taken ages to build up. Every- 



