THE GROWTH OF CFNES 



tassel-seed, t^ t^, he turned the male flowers into females (and at 

 the same time restored fertility to the female flowers). The segre- 

 gation T, — t^ in the silklcss stock then gave equal numbers of 

 males and females in the progeny crossed with t^ t^, as follows: 



Female 



Male 



t. r, Su s.. 



X 



r. L 



'k ^k 



t. L 



T. t 



s ^k ^k 



Thus a single gene segregating acts as a trigger or switch for sex 

 differentiation or determination. But it does so only against a 

 selected genetic background. In Jones' experiment this background 

 was maintained by keeping the silkless stock away from the ordinary 

 stock. In nature the same result could be attained correspondingly 

 by isolation, geographical or genetical, from the older stock. It 

 could also be attained, even where crossbreeding is not prevented, 

 by a suppression of recombination between the two genes arising in 

 any of the various ways we have seen. The two dioecious types and 

 the old monoecious type would then work as though expressing a 

 multiple allelomorphic series. 



In the fish Lehistes, with the simplest known mechanism, a single 

 gene pair (making an X — Y pair of chromosomes) does in fact 

 seem to act as the switch. Again it acts against a correct background 

 provided by the autosomes. The switch gene, as we saw, is asso- 

 ciated with additional polymorphic colour genes. It is a super-gene 

 in process of birth. But it is not yet stabilized, for it can still be 

 replaced by a new pair of genes in another pair of chromosomes. 

 Wingc has found two such alternative systems. Both of these 

 provide autosomal switch actions. One of them w^orks only in 

 combination with X Y. This type necessarily segregates X X and 

 Y Y types and the double segregation of course means that half 

 the progeny are wasted in each generation. The other wojks in 

 combination with X X and provides a stable new system. The X X 



340 



