344 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1958 
most characteristic feature of the Mendelian gene is that it never 
blends, but retains its identity and properties intact for long periods 
of time until it mutates, after which it remains intact in its new 
condition until it eventually mutates again. This means that the 
amount of variation, or variance, present in a population resulting 
from previous mutations, is not only conserved through generation 
after generation, but is actually increased as a result of the recom- 
binations of the gene complexes in their innumerable possible per- 
mutations. This power of increase is one of the most important 
results of the biparental method of reproduction and is the reason 
why organisms possessed of this mechanism have evolved further than 
those that lack it. 
This conservation of variance is to be considered in relation to the 
rate at which mutation normally occurs. It has been calculated that 
in organisms as diverse as a bacterium, a maize plant, a fruit fly, 
and in man, any given gene mutates in one in about half a million 
individuals. It is also clear that this rate is itself the result of 
selection, and that although seemingly slow, it has been adequate to 
provide the requisite basic heritable variation which the mechanism 
of germ-cell formation and fertilization has multiplied, and on which 
selection has worked to produce whatever evolution has taken place. 
In other words, mutation not only need not, but must not be more 
rapid than a slow rate. This rate is immeasurably slower than 
what it would have to be if “blending inheritance” were a fact, and 
Darwin’s difficulty in accounting for an adequate supply of variation 
is lightened by that amount. 
As the originating mechanism for basic heritable variation, muta- 
tion has naturally been intensively studied. It has been found that 
certain physical and chemical agents, including radioactivity, can ac- 
celerate the rate at which mutation would naturally occur, but that 
these induced mutations are similar to those which occur and recur 
normally, and no correlation whatever exists between the mutagenic 
agents and the quality or “direction” of the mutations. Mutations 
take place with “blindness and molar indeterminacy,” as H. J. Muller 
has expressed it. This isa finding of capital importance, for it shows 
that there is no basis for attempts to explain the origin of heritable 
variation by appealing to environmental factors to evoke appropriate 
responses, or to the internal factors to make such responses. Nor is 
there any basis for the view that the environment would evoke ap- 
propriate heritable responses if its actions were continued for a sufli- 
cient time, because, as J. B. S. Haldane showed, such responses as 
might be significant in evolution would be detected within the period 
of the experiment carried out. 
In organisms that reproduce by simple division of the whole body, 
such as bacteria, special conditions apply because reproduction in them 
