BALANCED LETHALS 339 



type of hybrid, and reciprocal crosses may be of different types or 

 the same types in different proportions. 



4. Some of the diploid mutants (e.g., CE. nihrinervis from CE. 

 Lamarckiana, De Vries, 1918) and crosses [e.g., CE. grandiflora x 

 CE. Hookeri, Cleland and Oehlkers, 1929) yield on self-fertilisation 

 a much higher proportion of new forms than their parents did ; 

 others resemble their parents in hereditary properties, and others 

 again resemble the Californian species, CE. Hookeri, in being 

 absolutely true-breeding and fertile : they have entirely lost their 

 parents' exceptional properties. 



The special system of inheritance found in GEnothera can be 

 expressed and understood in mendelian terms if we accept the 

 complex-heterozygote hypothesis of Renner (1917, cf. 1925) and the 

 balanced-lethal hypothesis of Muller (1918). The first assumes that 

 each of the mutating species of CEnothera is a hybrid that produces 

 two kinds of gametes. These gametes have genetic complexes which 

 are distinguished by differences more profound that those that are 

 inherited in a mendelian way in other organisms. They affect the 

 whole habit and structure of the plant, although they behave as a 

 unit in inheritance. In these respects they resemble the sex 

 difference in animals. 



The second hypothesis assumes that gametes of neither complex 

 can yield viable homozygotes. Each complex is " lethal " in 

 combination with an identical complex. In terms of Muller 's 

 observations on Drosophila (1917, cf. 1918), it may be said that each 

 complex of factors is linked with a recessive lethal factor with no 

 crossing-over between them, so that the heterozygote for two 

 complexes is viable and breeds true. The whole system can then 

 be described as depending on a " balanced lethal " mechanism 

 (Fig. 104). 



A third assumption can then be made, that dissociation through 

 crossing-over occasionally takes place (a) between a complex and its 

 lethal factors, and (b) between two parts of a complex. The first of 

 these changes will permit the segregation of complete homozygotes ; 

 the second will result in the segregation of plants homozygous for 

 part of the complex. The segregates will be mutants. 



Further observations of Renner's (1919, « and b, 1921) explain 



