362 PRINCIPLES OF ANIMAL BIOLOGY 



Operation of Natural Selection. — How natural selection is believed 

 to work may be best illustrated from the standpoint of the genes. In any 

 species in which a certain gene is becoming either more or less common, 

 evolution is occurring. Even if the gene in question is recessive and 

 even if it occurs only in heterozygotes so as never to produce a visible 

 effect, if this gene is present in a gradually increasing or decreasing num- 

 ber of individuals, the species is evolving. Now most genes produced by 

 mutation are recessive. They cannot at first affect the visible or phys- 

 iological properties of the individuals, for these organisms are hetero- 

 zygous. During this early period, pure chance must be responsible for 

 the fluctuation of the prevalence of the new gene, as described above 

 (page 359). Many a new gene is lost before it can become established. 

 Most new genes are thus lost. Evolution would be going on at a tremen- 

 dous rate in some species if even a majority of mutations succeeded. Out 

 of the large number that occur only a few happen to become numerous 

 enough to begin to show their effects in homozygotes. Then they are 

 on trial. They may confer some advantage on their possessors, such as 

 longer life, more rapid growth, or greater strength. If the advantage 

 is one that enables them to leave more descendants, that gene tends to 

 become more prevalent. If on the contrary the new gene is harmful, in 

 Guch a way that its possessor leaves fewer descendants, it is checked in 

 any increase which it might otherwise enjoy. Mere harmfulness cannot 

 eradicate a gene altogether if it is recessive, for it may continue to exist in 

 heterozygotes beyond the reach of natural selection; but a very harmful 

 gene cannot become much more abundant than the level at which hetero- 

 zygotes begin to meet and mate, thus producing homozygotes. A less 

 harmful gene may become slightly more abundant than this, so that some 

 homozygotes appear; but it cannot replace the alternative gene which is 

 superior to it. A neutral gene, one conferring neither advantage nor 

 disadvantage, is at the mercy of chance. As pointed out before (page 

 359), local races within a species may come into existence in different 

 parts of a range by this accidental method. 



A considerable degree of variability may thus exist in any species. 

 Partly it is observable, as in the distinguishable local races ; but much of i't 

 is hidden in heterozygotes. At any particvilar time it is to be expected 

 that a species will exhibit approximately that part of its genetic composi- 

 tion which is most favorable; and ''favorable" means conducive to large 

 numbers of descendants. This most advantageous group of genes would 

 be expected to show, because, if they were not expressed, natural selection 

 would gradually bring them to expression. If, under these circumstances, 

 the environment were to change in some respect, so that certain genes 

 increased and their alternates decreased in value, without much question 

 the species would change toward the genetic make-up that had acquired 



