490 INTRODUCTION TO EVOLUTION 



this way would not be favorable for evolutionary change, since random 

 breeding of large numbers of individuals results in population equilibrium, 

 resistant to change. Change in conditions, or increase in severity of natural 

 selection, acting on such a large population might result in some shift in 

 gene frequencies, changing them to a new equilibrium which would then 

 be maintained as long as the new conditions persisted (Wright, 1931 ). But 

 the change would be slow, and would be reversible whenever conditions 

 altered again. Thus, while some increased adaptation might be brought 

 about by this means, no considerable evolutionary change would be pro- 

 duced. 



Conversely, very small populations are not favorable for evolutionary 

 change. They are too likely to lose genes by chance, through the action of 

 genetic drift, thus reducing their reserves of hereditary variability and 

 consequently their possibilities for further change. Owing to genetic drift, 

 also, members of a very small population may all come to possess genetic 

 characteristics which are of no particular value or are even deleterious. 

 Thus observers of species inhabiting small oceanic islands find, on the one 

 hand, reduced variability among them and, on the other, possession of a 

 variety of seemingly nonadaptive characteristics. The disharmonic nature 

 of the fauna of oceanic islands (p. 283) frequently results in absence of 

 competitors for a given environmental niche. Therefore, relatively ill- 

 adapted animals can continue to exist — even animals so ill adapted that 

 they would not succeed in the stiffer competition characterizing life on 

 continents. Zimmerman (1948) recorded the unusual number of flightless 

 insects on the Hawaiian Islands and explained them as mutant forms which 

 can survive under conditions of reduced competition found on oceanic 

 islands. The influence of size of population on variability was clearly 

 demonstrated in the observation of an isolated colony of butterflies over a 

 prolonged period by Ford (1949). For the first few years specimens re- 

 mained rare; it was noted that those taken were very constant in appear- 

 ance. In later years the species increased greatly in numbers; "an extraordi- 

 nary outburst of variation took place while the numbers were rapidly 

 increasing, and many of the more extreme aberrations were deformed." 

 The author continued, "When the population became stabilized again at 

 the new and high value, uniformity was restored, yet the constant form 

 which was then established differed in appearance from that which existed 

 before the outburst of variability." 



Populations broken up into subpopulations provide a means by which 

 nature can make "experiments," to employ a figure of speech suggested by 



