APPENDIX B 365 



the levelling effects of bi-parental reproduction. Except in the 

 Mendelian sense such a strength of prepotency is unknown in 

 nature. 



De Yries, indeed, supposes that in "mutating species" many 

 individuals tend to change simultaneously in the same direction, 

 that they tend to be infertile with the parent type, and that thus 

 new " elementary species " are established. But probably all the 

 instances of epidemic mutation recorded by him are instances of 

 reversion in species which have undergone rapid evolution 

 reversions of the kind that is common enough in prize breeds of 

 animals and plants. It is noticeable, besides, that none of the 

 examples quoted by him imply increased adaptation. Most of 

 them are instances of unimportant changes in colour and shape 

 of leaves or flowers unimportant that is regards the adaptation 

 of the whole individual to the environment and of its parts to 

 one another. 



If evolution has resulted from the selection of mutations which, 

 unlike fluctuating variations, are stable and never disappear, 

 except through the action of selection, how shall we account for 

 the regression of the old-established parts in the total absence of 

 selection the vestigial parts of plants and animals for example ? 



For some thousands of years man, especially civilized man, 

 has been stringently selected by disease, and in a number of 

 instances has undergone great evolution against it. Obviously 

 human variations, as regards resisting power to disease, are of 

 a most fluctuating type. In the same community, even in the 

 same family of people of the most resistant race, we may find 

 individuals who are highly resistant to this or that disease, 

 others whose resisting power is very low, and all intermediate 

 types. The hypothesis of evolution by prepotent mutations is 

 excluded by the fact that, when a race (e. g. Spanish) which has 

 undergone evolution against a disease (e.g. tuberculosis) crosses 

 with a race (e.g. American Indian) which has undergone no 

 evolution, the offspring blend the characters of their parents, and 

 are neither so resistant as the one race nor so susceptible as the 

 other. Surely, if in this most familiar and easily investigated 

 example, evolution has proceeded on lines of "fluctuating" 

 variations, we have no right as scientific men to assume that 

 in all other instances it has proceeded on different lines. 



Plants are more variable than animals ; or rather species of 

 plants tend more to break up into a greater number of varieties 

 than species of animals. Parthogenetic species of plants (e. g. 

 hawks weed, dandelion) or self -fertilized species (e. g. wheat, pea) 

 possess an exceptionally large number of varieties. 1 The obvious 



1 See De Vries, Species and Varieties, pp. 59-61. "Thousands of forms 

 (of dandelion or hawksweed) may be cultivated side by side in the 

 Botanical Gardens, and exhibit slight but undoubted differentiating 

 features, and reproduce themselves truly by seed." 



