636 



ECOLOGY AND EVOLUTION 



nomenon referred to as neoteny. Examples 

 are found among salamanders such as the 

 aquatic Necturus, which has the caeno- 

 genetic gills of the ancestral larvae in the 

 neotenous form, the typical adult stage pre- 

 sumably having been lost. Substitute re- 

 pro ductives of termites have nymphal wing 

 buds and undeveloped eyes. The adult fe- 

 male glowworm (Phengodes) is larviform, 

 even though it passes through a semipupa 

 stage. The male of Phengodes is a normal 

 adult beetle. 



A wide variety of evolutionary modifica- 

 tions of developmental processes has been 

 summarized by de Beer (1940). The envi- 

 ronment at all stages of development has 

 exerted a profound evolutionary eflFect. An 

 adaptation fitting one stage to its particular 

 habitat may move into either earlier or later 

 ontogenetic stages during subsequent evolu- 

 tion. Man himself has apparently evolved 

 to a marked degree by moving embryonic 

 characters into the adult stage (Haldane, 

 1932, p. 149; Ariens Kappers, 1942), a 

 process referred to as foetalization or pae- 

 domorphosis (Gregory, 1946, p. 354), and 

 closely similar to neoteny. 



One can hardly discuss caenogenesis 

 without mentioning the contrasting devel- 

 opmental principle of palingenesis, usually 

 referred to as recapitulation. This concept 

 is discussed more fully under Regressive 

 Evolution (pp. 672, 673, 678). Caeno- 

 genesis may also be contrasted with deu- 

 terogenesis—ihe appearance of new adap- 

 tive characters toward the end of life 

 (Swinnerton, 1938). Such characters as the 

 wings of insects, which function only in the 

 adult stage, afford a good example of 

 deuterogenesis. Functional reproductive ad- 

 aptations are also characteristic of adults 

 only, but the advantage pertains to the new 

 generation. In contrast with the caenogene- 

 tio evolution of the sterile castes of ter- 

 mites, worker and soldier ants (Formi- 

 cidae) obviously are deuterogenetic. Social 

 adaptation in ants, through division of 

 labor (Fig. 253), is largely the result of 

 adult modifications, each ant caste having 

 developed by complete metamorphosis 

 through larval and pupal stages. 



The behavior of the young marsupial at 

 the time of birth exemplifies a combination 

 of adaptations (Matthews, 1944). The 

 young, after birth, crawls by means of well 

 developed forelegs into the mother's pouch, 



where it attaches itself to a nipple. This 

 behavior, together with the special devel- 

 opment of the foreUmbs and their muscu- 

 lar and nervous connections, is surely 

 caenogenetic, and the pouch and nipple are 

 surely deuterogenetic. The young has a 

 functional mesonephros at the time of birth, 

 a palingenetic pronephros during embryo- 

 logical development, and later develops a 

 metanephros that functions after the mes- 

 onephros has been reduced subsequent to 

 emergence from the pouch. 



Known evolutionary sequences are nearly 

 always based upon morphology. It has al- 

 ready been pointed out (p. 634) that evo- 

 lution of reflex and instinctive behavior fol- 

 lows the same principles as the evolution 

 of structure (Emerson, 1938; Hooker, 1944; 

 Figs. 231, 232, 233). One might expect 

 homologous, convergent, caenogenetic, deu- 

 terogenetic, and palingenetic behavior. 

 Some reservation is necessary, however, 

 since most behavior cannot be described in 

 terms precise enough to warrant the as- 

 sumption of homology. 



Humphrey (1944) describes neurons in 

 the embryonic human central nervous sys- 

 tem as homologous to the Rohon-Beard 

 cells of amphibians and lower vertebrates 

 that function as a temporary sensory ap- 

 paratus in the amphibian embryos and 

 larvae (Du Shane, 1938). These neurons 

 seem to be nonfunctional in man and are 

 replaced by functional intramedullary bi- 

 polar sensory cells that later develop into 

 unipolar cells. If these primitive neurons 

 are considered palingenetic, it is not diffi- 

 cult to imagine palingenetic behavior re- 

 sulting from such an inherited nerve pat- 

 tern. The rather futile action of dogs in 

 scratching dirt after defecation appears to 

 be an example of inherited behavior that 

 has undergone regressive evolution (Emer- 

 son, 1938, p. 280). 



If one grants that many characteristics 

 of living organisms are understood only 

 through knowledge of the functions of 

 homologous characters in ancestral forms, 

 and that correlations of ontogeny and phy- 

 logeny may be expected if genetic mech- 

 anisms underlie both, then many essential 

 features of the recapitulation theory may 

 still be accepted (pp. 677, 678). De Beer 

 (1940) takes the opposed position that 

 "phylogeny plays no causal part in deter- 

 mining ontogeny," and that it "does not 



