THE ORIGIN OF SPECIES 



misunderstood by his proponents as well as by his opponents. It has often 

 been said that he regards small, isolated populations as optimum for rapid 

 evolution. But this is not the case. Because genetic drift predominates over 

 selection in such populations, he believes that they will show a higher 

 degree of homozygosity than will more typical populations, and that they 

 will, on the whole, tend to be rather poorly adapted. As a result, they may 

 well become evolutionary blind alleys. 



A corollary to the Sewall Wright effect is what Stebbins has called the 

 "bottleneck" phenomenon. It is often said that the numbers of a species 

 tend to remain approximately constant in any locality. But every field 

 biologist knows that a species which is very abundant in one year may be 

 very difficult to find in another year, only to be followed by another in- 

 crease. In years of scarcity, small populations assume an especial impor- 

 tance, for they are the only source from which the species can again be 

 built up. Hence the term "bottleneck." Accidental changes in the genetic 

 make-up of such bottleneck populations will therefore determine changes 

 in the larger populations to be derived from them, and these changes will 

 generally be nonadaptive in character. For example, the arctic hare peri- 

 odically reaches a peak of population at which disease decimates the 

 population. There follows a "no rabbit year" and then several years of 

 recovery. Accidental changes in gene frequencies during the years of 

 scarcity must have a profound effect upon the populations of years of 

 abundance. The same thing must be true of the lemming, a small rodent 

 of the Scandinavian mountains. Every few years its population reaches 

 a prodigious level, and epidemics do great damage. The lemmings then 



FlGtTRE 100. AlLOMETRY OF THE 



"Tail" of the Swallowtail But- 

 terfly, Papilio dardanus. This mod- 

 erately positive allometry is shown 

 only by the males. (From Huxley, 

 "Problems of Relative (Growth," Meth- 

 uen & Co., Ltd., London, 1932. ) 



260 



