HUMAN POPULATION IN THE FUTURE. 505 



other races in having a power '.f artificially counteracting 

 external changes, yet there are limits to this power; and, 

 even were there no limits, the changes could not fail to 

 work their effects indirectly, if not directly. If, as is thought 

 probable, these astronomic cycles entail recurrent glacial pe- 

 riods in each hemisphere, then, parts of the Earth that are at 

 one time thickly peopled, will at another time, be almost de- 

 serted, and vice cersd. The geologically-caused alterations of 

 climate and surface, must produce further slow re-distributions 

 of population ; and other currents of people, to and from different 

 regions, will be necessitated by the rise of successive centres 

 of higher civilization. Consequently, mankind cannot but 

 continue to undergo changes of environment, physical and 

 moral, analogous to those which they have thus far been 

 undergoing. Such changes may eventually become slower 

 and less marked ; but they can never cease. And if they can 

 never cease, there can never arise a perfect adaptation of 

 human nature to its conditions of existence. To establish 

 that complete correspondence between inner and outer actions 

 which constitutes the highest life and greatest power of self- 

 preservation, there must be a prolonged converse between the 

 organism and circumstances that remain the same. If the 

 external relations are being altered while the internal rela- 

 tions are being adjusted to them, the adjustment can never 

 become exact. And in the absence of exact adjustment, 

 there cannot exist that theoretically-highest power of self- 

 preservation with which there would co-exist the theoretically- 

 lowest power of race-production. 



Hence though the number of premature deaths may ul- 

 timately become very small, it can never become, so smal] 

 as to allow the average number of offspring from each pail 

 to fall so low as two. Some average number between two 

 and three may be inferred as the limit a number, however, 

 that is not likely to be quite constant, but may be ex- 

 pected at one time to increase somewhat and afterwards 

 to decrease somewhat, according as variations in phyeica] 



