The Amphiploids 305 



possible now that many fertile amphiploids can he produced, will 

 not face the same difficidties confronting a breeder who tries to com- 

 bine the characters of the already \vell-kn()\vn Upland and Sea Island 

 cottons. 



If some chromosomal mechanism prevents the recombinations of 

 genes contributed by each parent, then merely growing large prog- 

 enies and exercising selection can hardly be expected to yield re- 

 sults.^'"'' The evolution of the tetraploid from dij)loids can be ex- 

 plained by the hybridization and doubling of chromosomes. This 

 does not explain the difterentiatirjn of the tetraploid species after 

 they once originated as an amphiploid. An argimient supported by 

 considerable data^*"' asserts that a structural differentiation of chromo- 

 somes was basic to speciation and this was of the cryptic type, that is, 

 in very small segments, so that a differentiation could not be ob- 

 served by pairing or irregularly arranged chromosomes at meiotic 

 metaj)hase. Therefore, a genetic hybridity and a hybridity caused by 

 the differentiation of small chromosomal segments could not be de- 

 tected by the ordinary genetic and cytological methods. The nature 

 and extent of chromosomal differentiation may be measmed by trac- 

 ing marked genes in subsecjuent generations and recording the rates 

 at which the genes are lost by successive backcrossing. Such chromo- 

 somal differentiation may be important in Gossypirim.'^^^ At least, 

 the suggestion has led to inflection on these problems in polyploidy. 



Among the second generations of the interspecific hybrid between 

 G. hirsiitum and G. barbadense, asynaptic genes account for the ste- 

 rility observed, notably when certain parents are used." Genes for 

 asynapsis have been foimd in both genomes A and D. By the use of 

 trisomies, additional data about these asynaptic types have been col- 

 lected. The fully sterile plants eliminate the completely asynaptic 

 types, but partial asynaptic types are carried along.^^ Some of the 

 j)hen()mena attributed to a cryptic structinal hybridity might be ex- 

 ])lained on the basis of asynaptic and partially asynaj)tic genes. ^''' 



Sterility resulting from asynaptic genes is a kind of genic-*^ sterility 

 and may well be important in such sterility that causes failure in 

 chromosomal pairing. The extreme sterility at the diploid hybrid 

 level can be overcome by doubling the chromosomes. But a sterility 

 due to asynaptic genes is not cmed through doubling the nimiber of 

 chromosomes. Later generations introduce new problems in maintain- 

 ing the fertility level as well as the characters brought together in the 

 hybrid. If by selection some desirable characters contribtited into 

 the hybrid are eliminated and inidesirable ones retained, polyploid 

 breeding is faced with a difficult task. To incorjx)rate into commercial 

 varieties the desirable characters foimd in other sj^ccies can be ])ut 



