496 EVOLUTION, GENETICS, AND EUGENICS 



make these genes differ from each other in the first place. It should in 

 addition be noted that changes due merely to differences in the gross 

 proportions of entire groups of genes must be relatively incapable of 

 that delicate adjustment which is required for evolutionary adaptation. 

 And as to the question, frequently raised, whether all evolution is ulti- 

 mately due to mutation, this is necessarily answered in the affirma- 

 tive by our definitions of the gene and of mutation, which designate the 

 gene as any unit of heredity, and mutation as any transmissible change 

 occurring in the gene. The question of the basic mechanism of evolu- 

 tion thus becomes transferred to the problem of the character, fre- 

 quency, and mode of occurrence of mutation, taken in this precise, 

 yet comprehensive sense. And since eugenics is a special branch of 

 evolutionary science it must be equally concerned with this problem. 

 In choosing the body of data wherewith to attack these questions of 

 mutation, in their new form, it must imfortunately be recognized that 

 the results with the evening primrose, Oenothera, although they formed 

 the backbone of the earlier mutation theory, can no longer be regarded 

 as having a direct bearing on the modern problem, since they cannot 

 be shown to be due directly to changes in the genes. Certain of them, 

 such as gigas, lata, scintillans, etc., have been proved by Geerts, Lutz, 

 Gates, and others, to be due to abnormaUties in the apportionment of 

 the chromosomes. Very valuable information on the genetics of cases 

 of this sort is now being obtained, especially in the work of Blakeslee, 

 Belling, and Farnham on much clearer cases of similar character in the 

 Jimson weed, and, finally, in the work of Bridges on the fruit fly Droso- 

 phila. Most of the other so-called mutations in the evening primrose 

 appear to be due to the normal hereditary processes of segregation and 

 crossing over, working on a genetic constitution of a special type. Evi- 

 dence for this was obtained in my analysis of the analogous case exist- 

 ing in the fly Drosophila, as follows. It had previously been shown by 

 de Vries, and further elaborated by Renner, that germ cells or indi- 

 viduals of Oenothera bearing certain genes always died, in such a way 

 that all the surviving individuals were heterozygous (hybrid) in regard 

 to these genes. I later showed, through work on Drosophila, that 

 when such a condition (there called "balanced lethal factors") exists, 

 the situation tends to become still further compHcated through the 

 presence of other heterozygous genes, which are linked to those which 

 cause death. When one or a group of these non-lethal genes crosses 

 over (separates) from the lethals, as they occasionally do, they may 

 oecome homozygous, producing a visible effect. Thus new types of 

 individuals appear which may be ascribed to "mutatiop" whereas thej 



