The Experimental Polyploids 285 



new s[)ecies such as the Cucurhild inosdiata X C. maxima amphiploid 

 combines good traits from two diploids. A new species of economic 

 potential is apj:)arent. However, intersj^ecific segregations in tlie fifth 

 and sixth generations show that a lack of Lniiformity can be expected 

 (cf. Chapter 12) . Such variation is not what the breeder hopes for 

 in a true l^reeding variety. By transfer of whole genomes into a hy- 

 brid the characters of the polyploid can be influenced. If in later 

 generations there is pairing between the two genomes that originated 

 \\\x\\ the two species, the chance for segregation is good. If the segre- 

 gates are undesirable and if the interchange is so great that the 

 original type is lost, all the transfer is circumvented by the after- 

 breeding effects. Transfer in Gossypiimi has presented a very difficult 

 problem, that of introducing the good characters and maintaining all 

 the original traits of the cultivated varieties. In spite of the ])roblems, 

 the principle of transfer is basic in polyploid breeding.*""' 



The advantages balanced against the disadvantages are necessary 

 for a final evaluation. •''i No tetraploid within a certain species may 

 be expected to surpass the diploid in all respects. Therefore, the 

 desirable traits balanced against the unfavorable ones should be cal- 

 culated to see whether the new result is in favoi of the tetraploid or 

 the diploid. Triploid sugar beets are not perfect, but there is the 

 important fact that the triploids can be grown to a larger root size 

 before the percentage of sucrose decreases than is the case for the 

 diploids."**' In this way the triploid has an advantage over the dip- 

 loid, Avhile for seed production, germination, and growth problems 

 the triploid is sometimes at considerable disadvantage beside the 

 diploid. Tetraploid rye offers another notable example of balancing 

 two sets of characters. ^^ 



All plants arising from treated generations may not be totally 

 tetrajjloid. The diploid cells may be found mixed with the tetra- 

 ploid, and a mixoploid condition may persist.^" Or the layers of cells 

 may differ one from the other, so that the shoot apex is stratified with 

 respect to its ploidy.-'^ These are called periclinal chimeras discussed 

 in Chapter 14 (The Aneuploids) .i"' From the point of view of poly- 

 ploid breeding the mixoploids and chimeras are very important prob- 

 lems. The reversion of jjolyploid to diploid is sometimes explainable 

 on the basis of a chimera, or sometimes it may arise from cross-breed- 

 ing. 



Stabilizing the polyploid by selection and In preventing the re- 

 version to the diploid or through segregation, to some inferior type 

 is a problem that confronts the plant breeder after the polyploid has 

 been produced. The first and second generations may be quite uni- 

 form, but later generations less so. Or the first generation may have 

 defects that yield to selection in later generations. The effectiveness 



