REMARKS ON THE EXPLANATION OF HETEROSIS 



CORRADO GINI 



Rome, Italy 



It is well known that the phenomenon of heterosis, or luxuriance, of hy- 

 brids, which occirs more especially in hybrids of the first generation, is 

 generally explained by the Mendelian theory, and by the hypothesis that 

 the factors of the favorable characters (or merely the more energetic factors) 

 are more or less completely dominant, and the factors of the unfavorable 

 characters (or the less energetic factors) recessive. Considering that, as far 

 as the factors determining the differential characters of the parents are con- 

 cerned, the hybrids of the first generation are heterozygous, it follows that 

 in the hybrids of the first generation we shall have for each character the 

 manifestation of the more favorable or more energetic factors. The con- 

 sequent heterosis will therefore be attenuated in the second generation to 

 an extent in keeping with the reduction of the percentage of the heterozygous 

 individuals. 



So far the theory is consistent. But it ceases to be so when it goes on to 

 assert that the attenuation of heterosis should be progressively intensified in 

 the successive generations. 



An example will serve to illustrate this point. In the example taken I will 

 suppose that the character is determined by several factors, of which some 

 at least are contained in the same chromosome, and are therefore linked to- 

 gether in heredity, in conformity with the hypothesis to which it has been 

 considered advisable to have recourse in the latest statement of the theory 

 under examination. For the sake of simplicity I will only consider the case 

 in which the character factors are two, and both contained in a single chro- 

 mosome, but the argument holds good if the factors considered are more 

 numerous and if, instead of being contained in one chromosome, they are 

 distributed over several chromosomes, which, according to the Mendelian 

 theory, segregate one independently from the other. 



Let us, then, take an individual containing in the chromosome under con- 

 sideration, two factors which we will designate by the figures 1 and 3, in a 

 homozygous state; and let us suppose that it crossed with another individual 

 whose chromosome, also in a homozygous state, contains, instead, the fac- 

 tors 2 and 4. Let us further suppose that factors 1 and 2 determine in the 



421 



