236 AN INTRODUCTION TO MODERN GENETICS 



and allowing no crossing-over between this part and its partner in the 

 Y. This complete suppression can only be caused by the genetic control 

 of the position of chiasma formation. The non-crossing-over part of 

 the Y presumably originally carried hypomorphic allelomorphs of 

 the factors in the X. We find this condition in frogs, where the F 

 factors in the X and the / in the Y are nearly equal in strength in those 

 races with a considerable hermaphrodite period. In more highly 

 specialized types, a clearer decision between male and female is 

 obtained by diminishing the importance of the Y allelomorphs. This 

 comes about as a natural consequence of the fact that the Y is never 

 homozygous, so that hypomorphic mutations (and most new mutations 

 are hypomorphic) will always be masked and cannot be eliminated 

 from it by adverse selection, as they would be in any other region of 

 the chromosomes. Similarly the Y will lose any factors it may have 

 originally have had tending to produce the heterogametic sex and the 

 autosomes will take over the function of balancing against the Xs. 

 Thus the Y eventually becomes empty of genes and may disappear. 



4. The Meaning of ^^ Sex" 



The discussion has now reached a point where we can try to define 

 what seem to be the most important concepts. In the first place, the 

 important concept from the functional point of view is that of sexual 

 reproduction in the sense of reproduction by single cells in such a way 

 as to encourage recombinations of factors. In its fully developed form, 

 this involves the mechanism of fertilization following on reduction of 

 chromosome number by meiosis, with consequent segregation. There 

 are, however, less efficient methods, but those known to us can prob- 

 ably be regarded as degenerate types derived from fully sexual repro- 

 duction. One form of degeneration is found in diploid apomixis by the 

 production of restitution nuclei ; a slight amount of recombination is 

 still possible by crossing-over in the four strand stage (p. 61). Another 

 degeneration is complementary to this; the suppression of crossing- 

 over with the retention of chromosome segregation, as happens for the 

 Y chromosome or the autosomes of male Diptera. 



From a morphological point of view, we can define sexual differen- 

 tiation in its broadest sense as anything which tends to increase the 

 chance of crossing of different individuals. The type of differentiation 

 involved is very different in different organisms. It is the result of the 

 activation, by a differential factor or group of factors, of one or the 

 other of two alternative reaction systems of the genotype. These 



