602 



ECOLOGY AND EVOLUTION 



cytoplasmic factor (Sonneborn, 1945a, 

 1948). 



Through experiments on the ability of 

 certain yeasts (some undergoing cell divi- 

 sion and some not) to acquire the enzy- 

 matic apparatus necessary to ferment galac- 

 tose, Spiegelman and Lindegren (1944) 

 conclude that such an adjustment may re- 

 sult from (1) the natural selection of ex- 

 istent variants with the desired characteris- 

 tics from a heterogeneous population; (2) 

 induction of a new enzyme by the sub- 

 strate in all the members of a homogeneous 

 population, resulting in an increase in the 

 measured enzymatic activity of the popula- 

 tion; (3) a combination of natural selec- 

 tion and the action of an induction mecha- 

 nism on those selected (also see Spiegel- 

 man, Lindegren, and Lindegren, 1945; 

 Spiegelman and Kamen, 1946). 



Darlington (1944) postulates separate 

 systems of hereditary mechanisms: (1) 

 nuclear system (genes in chromosomes), 

 (2) corpuscular system (plastids or plasto- 

 genes in green plants), and (3) molecular 

 system (plasmagenes in cytoplasm with no 

 visible bodies and chemically variable in 

 degree of equilibrium). Darlington's nu- 

 clear system might well be divided into 

 (a) gene and (b) chromosome mecha- 

 nisms. Chromosomal aberrations and poly- 

 ploidy without gene change produce hered- 

 itary modification. There may be interac- 

 tion between all these systems in the cell 

 (Sonneborn, 1945, 1945a). 



Plasmagenes have striking similarity to 

 viruses (Altenburg, 1946; Sonneborn, 

 1948). Both viruses and plasmagenes are 

 proteins that require desoxyribose nucleic 

 acid in their composition. Some induced 

 mutations that react adaptively to the in- 

 ducing agent may involve plasmagenes 

 rather than nuclear genes (Spiegelman, 

 Lindegren, and Lindegren, 1945). 



Cytoplasmic characters, induced by 

 either gene action or environmental agents, 

 may persist through several generations, 

 but gradually disappear because of a lack 

 of persistent autocatalysis. This phenom- 

 enon has been called Daiiermodifikation 

 (JoUos, 1932; Taliaferro and HuflF, 1940; 

 Hoare, 1943; Caspari, 1948). Although this 

 "cytoplasmic lag" has been interpreted as 

 an indication of Lamarckian inheritance, 

 especially by earlier authors, Sonneborn 

 (1942) indicates that at least some cases 



may be explained by MendeUan inheritance 

 together with an interaction of the genes 

 with cytoplasmic factors. Genes are always 

 autocatalytic. Autocatalysis may be found 

 in cytoplasmic systems, where cytoplasmic 

 inheritance through plastogenes or plasma- 

 genes is possible. In either case, induced 

 modifications by the environment, even 

 when inherited and adapted to the stimulus, 

 are not strictly Lamarckian and had best 

 not be used to justify Lamarckian theories 

 of the inheritance of acquired somatic 

 characters. 



GENETIC VARIATION IN RELATION TO 

 POPULATION NUMBERS, POPULATION 

 RHYTHMS AND DISPERSION 



"The elementary evolutionary process is 

 . . . change of gene (or chromosome) fre- 

 quency" (Wright, 1945, p. 415). A species 

 divided into many partially isolated local 

 populations (p. 603), each with a tendency 

 toward fixation of a uniquely balanced 

 gene pattern, provides an enormous poten- 

 tial variability through occasional crosses 

 between populations. Selection may oper- 

 ate upon the group as a unit (p. 683). 



When a partial discontinuity results from 

 extrinsic causes, genetic recombinations be- 

 tween groups may be expected to be large- 

 ly confined to a narrow zone of popula- 

 tion overlap (Huxley, 1939a). As a large 

 statistical generalization, a sparse popula- 

 tion between two areas of greater density 

 provides a hindrance to free gene flow, 

 which wdll enable fixation, mutation, and 

 selection to build up distinctive gene pat- 

 terns in the two dense areas (p. 611). The 

 extension of the ranges of these more har- 

 monious and viable dense populations will 

 tend to narrow the intermediate zone of 

 intergradation (p. 623). 



Species with large numbers are, in gen- 

 eral, more variable than those with small 

 numbers (Fisher, 1937; Mather, 1943), 

 for with the smaller numbers inbreeding 

 tends to increase homozygosity. There are 

 critical points in the breeding population 

 size above which httle increase in varia- 

 biUty occurs (Simpson, 1944, p. 67). In 

 small populations the natural accidents of 

 sampling result in random fixation of muta- 

 tions and gene frequencies (Fig. 229). 

 Small interbreeding populations may result 

 from a drastic reduction in numbers fol- 

 lowing severe climatic, predatory, or epi- 



