NATURAL SELECTION 



679 



olds are under the influence of modifying 

 genes during evolution. Hence, we may ex- 

 pect to find eyed and pigmented salaman- 

 ders that do not become bUnd or pale in 

 the absence of fight, bfind and pale cave 

 salamanders with all the requisite genes for 

 normal eye and pigment development, and 

 eyeless white salamanders that breed true 

 in perpetual fight. 



Leach (1944) suggests that the loss of 

 the forebrain during regressive evolution of 

 amphioxus has resulted in the loss of the 

 pituitary gland with its thyrotropic hor- 

 mones. The thyroid (endostyle) is thus 

 supposed to have degenerated and carried 

 with it other systems dependent upon it. 

 In criticism of this theory, it should be em- 

 phasized that endocrine efiFects in the in- 

 vertebrate chordates are poorly understood, 

 and it may well be that they have not 

 evolved so far as in the vertebrates. 

 Secondly, if the endocrines had as impor- 

 tant general functions as they show in the 

 higher vertebrates, it is unlikely that they 

 would degenerate, even though the brain 

 and head were otherwise useless. Am- 

 phioxus probably evolved from an ancestor 

 with a much better developed head, but, 

 in becoming adapted to a sedentary sand- 

 burrowing fife, the selection pressure for 

 maintaining head structures was possibly 

 lessened. The head could have undergone 

 regressive evolution as new adaptations 

 arose, but only if the survival value of its 

 other functions were not so great as that of 

 the new balanced system as a whole. Selec- 

 tion is surely sorting organisms in terms of 

 developmental processes, physiological in- 

 tegration, and ecological adaptations. 



Dollo's so-called "law of the irreversi- 

 bility of evolution" (see W. K. Gregory, 

 1936; Huxley, 1942, p. 503; Cave and 

 Haines, 1944) may apply to the loss of 

 complex adaptations. Reversible mutations 

 certainly occur, but it would usually re- 

 quire too many within a balanced system 

 to have a genetically similar structure 

 evolve, once it has regressed. Fish gills do 

 not reappear in whales, even though the 

 embryo recapitulates gill clefts. The gill of 

 the primitive aquatic arthropod does not 

 reappear in the aquatic insect after a long 

 ancestry on land. We find, instead, a con- 

 vergent adaptation undoubtedly with a 

 different genetic background (Muller, 

 1939). 



If sfight genetic inliibitions prevent the 

 growth of a structure that has been lost, it 

 is possible that modification of these inhib- 

 iting genes might allow the character to re- 

 appear. For example, Wright (1934a) 

 raised heterozygous guinea pigs that devel- 

 oped the primitive pentadactyl foot. Castle 

 produced true-breeding pentadactyl guinea 

 pigs in 1906 that were apparently atavistic. 

 Robb (1937) discusses cases of apparent 

 atavistic appearance of the side toes in 

 modern horses. 



There is no more reason to look upon the 

 degeneration of locomotor and sensory or- 

 gans in a parasite as something to be de- 

 cried than to regret the evolution of ani- 

 mals from a chlorophyl-bearing ancestor, 

 with the consequent loss of the beautiful ca- 

 pacity to carry on photosynthesis (Free- 

 man, 1937; see p. 254). Parasites are de- 

 pendent upon their hosts, animals are de- 

 pendent upon plants, green plants are 

 dependent upon direct radiant energy. Re- 

 gressive evolution does not lead to extinc- 

 tion any more than does progressive 

 evolution. It may lead to dependence and 

 necessary integration with more complete 

 biological systems (p. 695). So does pro- 

 gressive evolution. The scientist has no vafid 

 reasons for an emotional attitude toward 

 these evolutionary tendencies that move the 

 organism in the direction of harmonious ad- 

 justment to a changing environment. 



In summary, regressive evolution is a 

 universal phenomenon affecting many for- 

 mer adaptive characters of organisms and 

 organismic systems. It is the result of prin- 

 ciples that apply also to progressive evolu- 

 tion, and the explanation of the process 

 gives us a better perspective on many as- 

 pects of natural selection. 



RETARDATION OF EVOLUTION 



Considerable differences in evolutionary 

 rate occur among different organisms, and 

 an understanding of the factors involved in 

 retardation should explain the existence of 

 primitive reficts ("fiving fossils") and 

 should also place the causes of evolutionary 

 change in sharper relief. 



Simpson (1944, p. 144) follows Hand- 

 lirsch in classifying reficts as (1) numeri- 

 cal (groups once abundant and now rare), 

 (2) geographic (groups once widespread 

 and now geographically restricted), (3) 

 phylogenetic (ancient groups exhibiting 



