THE MODIFICATIONS OF THE CHROMOSOME CYCLE 73 



homogenetic association, and even then it is probably never more 

 frequent than would be expected on a basis of pure chance, which gives 

 a gametic output from AABB of i AA : 4 AB : 1 BB. 



Polyploids of higher number than tetraploids behave in essentially 

 the same way, showing failure of association of homologues as a conse- 

 quence of an insufficient amount of chiasma formation, and homo- or 

 hetero-genetic association according to the differentiation between the 

 chromosome sets. 



3. Auto- and Allo-syndesis 



Homo- and hetero-genetic association are frequentiy discussed in 

 terms of auto- and allo-syndesis. The usages of these words are, how- 

 ever, by no means uniform, and even authors as closely associated as 

 DarUngton and Philp and Huskins use them in quite different senses. 

 Philp and Huskins use autosyndesis to mean the pairing of chromosomes 

 which are phylogenetically derived from the same diploid species; that 

 is, in the sense in which homogenetic is used above. Darlington, on the 

 other hand, uses autosyndesis to mean the pairing of chromosomes 

 derived from the same parental gamete. Thus the association of A 

 with A and B with 5 in a tetraploid AABB is autosyndesis for Philp 

 and Huskins, but allosyndesis for Darlington. 



Darlington's usage is in accordance with the original definition of 

 Ljungdahl,^ and when the word is used in this work it will be in the 

 sense he defined. The words homo- and hetero-genetic association, or 

 Philp and Huskin's auto- and allo-syndesis, can only be used if one 

 knows or can deduce the phylogenetic relations of the chromosomes, 

 but Ljungdahl's words may be used without implying any such specula- 

 tion. Thus if complete association takes place in a hybrid between a 

 diploid and a hexaploid, some of the chromosomes derived from the 

 latter must have associated, and autosyndesis must have occurred. This 

 fact may or may not enable one to make deductions as to the phylo- 

 genetic relations concerned, and thus to translate the facts into terms 

 of homo- and hetero-genetic associations. 



4. Secondary Pairing 



Homologous bivalents in an allopolyploid, even if not associated by 

 chiasmata, often lie near together in meiotic metaphase. This secondary 

 pairing is probably an expression of an attraction due to homology 

 balancing the normal repulsion between paired chromosomes. It may 



^ Ljungdahl 1924. 



