THE GROWTH or genes 



ot a haploid set oi seven chromosomes. Between the two com- 

 plementary super-genes of one species no crossing-over can take 

 place without giving rise to a mutant with smaller complexes and 

 so pulling the system to pieces. We therefore have the remark- 

 able situation in which all adjustment of each super-gene must be 

 relative to its partner. If, in Oe. muricata, the rigens complex becomes 

 less viable or effective as a pollen genotype, then adjustment is 

 restored and seed fertility is increased by the curvans complex 

 becoming less viable as an embryo-sac genotype. And this is what 

 has been happening, as shown in the Reimer effect, whereby rigens 

 spores replace their partners as embryo-sacs. With such a limited 

 recombination as is possible in a ring of 14 chromosomes it 

 may well be asked how such an elaborate adjustment can come 

 about. The answer is that it probably takes place through the 

 expansion of the complex, break by break, at the expense of the 

 distal segments of the chromosomes in which crossing-over can 

 still take place. In other words by a growth of the super-gene. 



The limit to the growth of the super-gene is reached only when 

 it comprises the whole of the nucleus. This may happen in two 

 ways. 



The first is in the fungal heterocaryon. This arises in heterothallic 

 fungi where an illegitimate fusion has occurred between the multi- 

 nucleate cells of hyphae of like mating type. As a result each cell 

 contains a number of different nuclei (Fig. 86). Each nucleus divides 

 by mitosis; but division is not co-ordinated, so that different cells 

 at one time, and the same cell at different times, have different 

 numbers of nuclei present in different proportions. Presumably the 

 need for a nucleus of a particular kind stimulates its division relative 

 to that of others. The nuclei, being incompatible, never fuse and 

 there is no meiosis. All the recombinations effected mechanically 

 between genes in ordinary sexual reproduction, are here effected 

 physiologically between nuclei by the changing chemical equilibrium 

 dependent on the changing reactions of nuclei, cytoplasm and 

 environment. The gene is still the unit of mutation, as we saw in 

 Neurospora, but if all but one of the mating types has been lost 

 so that the heterocaryon is obligatory the nucleus becomes the unit 

 of transmission. Each nucleus has become a single super-gene. 

 Inasmuch as this super-gene is under general cell control, it is 



342 



