CHOICE OF PARTNER 199 



Since the chromosomes in polyploids behave at zygotene as though 

 they consisted of a small number of " pairing blocks " each of these 

 changes of homology will prevent pairing in the block affected, 

 while pairing between identical chromosomes will proceed un- 

 obstructed. Thus the existence of limited numbers of effective 

 pairing blocks means that similar parts of dissimilar chromosomes, 

 if they are short, do not pair at random in a polyploid. There is 

 competition in pairing : hence the differential affinity of dissimilar 

 chromosomes in the presence and in the absence of identical partners 

 (D., 1928). 



An analogous observation has been made on a triploid hybrid 

 containing two sets of Crepis capillaris and one of Crepis tectorum 

 (Hollingshead, 1930 b). The tectorum chromosomes are regularly 

 left unpaired, although they pair with the capillaris chromosomes 

 in the diploid hybrid. Similarly the two Y chromosomes of Rumex 

 Acetosa regularly pair with the corresponding parts of the 

 X chromosome in the diploid, but in the triploid which contains 

 X X Yi Y2 they are left unpaired (Ono, 1928). The two X's are 

 completely identical and pair to the exclusion of the Y's, which 

 may be inferred from this to be identical with them only in small 

 segments (Ch. IX). The most striking examples of differential 

 affinity are shown by the pairing of chromosomes in the haploid 

 progeny of true diploids which have perfectly regular pairing. This 

 will be described in another connection (Ch. VIII). 



Differential affinity is characteristic of the behaviour of 

 homologous chromosomes of different sets in allopolyploid species, 

 as compared with that in their " haploid," " triploid," and 

 hybrid derivatives. In these derivatives, chromosomes derived 

 from the same parental gamete pair with one another, although 

 such pairing rarely occurs in the normal " homozygous " polyploid 

 species, where each has an identical mate. In these circumstances 

 the change in conditions seen in the doubling of the diploid Primula 

 kewensis is reversed, and it is possible to understand the relationship 

 that the natural polyploid has to the product of experiment such as 

 Primula kewensis. 



This " internal pairing " is known as autosyndesis and is opposed to 

 allosyndesis, the pairing of homologous chromosomes derived from 



