DARWIN-WALLACE CENTENARY—DE BEER 301 
split up into very small isolated colonies, changes in the relative fre- 
quencies of different genes might result from the errors of random 
sampling in the formation of the germ cells and their fertilization, 
without involving selection. This concept, advanced by Sewall 
Wright and known as “random genetic drift,” has been invoked as a 
possible cause of nonselective, nonadaptive evolution. It has, how- 
ever, been invalidated by the results of experimental studies in the 
field such as those of Fisher and Ford on moths, which have shown that 
selective factors are much more important than casual nonadaptive 
factors in determining the relative frequency of genes and in bringing 
about close adaptation to local environmental conditions. Even in 
comparatively numerous populations, from one generation to the 
next there are fluctuations in gene ratio larger than can be attributed 
to random sampling and which are controlled by selection. Such 
effects as may be due to random sampling in small populations can 
only be of negligible significance in evolution. 
Selection frequently works on a basis of compromise. Among the 
natives of Africa there is a condition known as sickle-cell anemia, in 
which the red blood corpuscles are deformed and shaped like the 
blades of sickles. This is controlled by a Mendelian gene which, when 
inherited from both parents (homozygous), produces an extreme effect 
which frequently kills the subject by thrombosis. When inherited 
from only one parent (heterozygous), the danger from thrombosis is 
not so great. In areas where malaria is present, however, there is a 
positive advantage in possessing the sickle-cell gene, because the 
malaria parasite cannot enter the abnormal red blood corpuscles. In 
accordance with the prevalence of malaria in the environment, there- 
fore, a balance is automatically struck in the population between the 
danger of dying from malaria if the individual has no sickle-cell gene, 
and the danger of dying of thrombosis if the individual has two sickle- 
cell genes. Survival value and ability to leave more offspring there- 
fore accrue to the possessors of one sickle-cell gene up to a certain 
frequency, and this example shows in what unexpected ways selection 
is able to make the best even of a bad gene complex. 
NATURAL SELECTION AND PALEONTOLOGY 
The paleontological record provides the evidence of the course which 
evolution has followed in the past. The fossil material is in places 
now so rich that it can be used for quantitative studies in evolution. 
First, the radioactive time clocks enable various levels of evolutionary 
lineage to be dated and the time measured during which certain 
changes have occurred. This provides quantitative evidence of evolu- 
tion rates. From such data estimates can be obtained of the duration 
times of genera and species. Statistical study of large samples of 
fossil materials enables the variability of the different species to be 
