TYING GENES TOGETHER 



pleiotropic one; not a single new effect but a mechanical union 

 of old ones. 



Mechanical union can be achieved in two ways which we have 

 discussed separately. The first is by localization of crossing-over, 

 either terminally or near the centromere. This method ties up large 

 segments of all the longer chromosomes : it ties them up perhaps 

 for good and at least for so long as the character of the race persists. 

 The second method is by inversions or interchanges. This is effective 

 only in the structural heterozygotes (where alone it is important) 

 and only for the segment concerned. In other words it is a specific 

 pegging device. In all species inversions are continually occurring. 

 They are of all lengths and in all parts of all chromosomes. They 

 are therefore always available for the task of super-gene construction. 



The flexibility of inversions as the building materials of super- 

 genes is well shown by Drosophila pseudoobscura. The third chromo- 

 some exists, as we saw, in a number of structural forms differing 

 by inversions. Some flies are heterozygous for sequences differing 

 by several inversions in this chromosome alone. Within these 

 inversions all effective recombination is suppressed in the female, 

 just as much as it is in the male, where there is no crossing-over 

 in any case. For such flies the whole third chromosome is thus a 

 single super-gene. Other flics again are hybrid for fewer or smaller 

 inversions, and some for none at all, so that every gradation can 

 occur within one group of individuals in the unit of recombination. 



If there were a dominant and visible gene to mark each of these 

 inversions, or characteristic sequences, in the fly, the population 

 would present us with the appearance of a polymorphic species with 

 a stable (or in some cases changing) equilibrium in the frequencies 

 of the different phenotypes. Such a situation has, in fact, been 

 revealed in a significant series of species of grouse-locusts by the 

 work of Nabours. The first, Acridium arenosum, is polymorphic in 

 regard to colours and patterns controlled by some thirteen genes 

 which recombine easily although in the same chromosome. The 

 second is Apotettix eurycephalus, in which corresponding genes show 

 about 7 per cent crossing-over between two tightly linked groups. 

 The last is Paratettix texanus with a complete suppression of crossing- 

 over between 24 out of the 25 colour pattern genes, which, however, 

 remain individually recognizable both by their correspondence with 



335 



