284 Colchicine 



ploid wheat combines three genomes, and after this process the optimal 

 number of 42 seems to be attained. 



Cross-fertilizing, or allogamous, species are more promising as a 

 group than the self-fertili/ing types. This general rule seems to hold 

 for a large number of plants included in the Svalof experiments. 

 Some qualification needs to be made, for the sampling was not as 

 extensive as might be desired. The changes from incompatability to 

 compatibility upon doubling the number of chromosomes is an in- 

 volved genetic problem, not merely a result of the tetraploid nature, 

 but consisting of a combination of events that create the changes.^'' 



1 he autoploids are almost without exception less fertile than the 

 diploids.***^ Therefore, seed and fruit yields, if dependent upon seed 

 production, will at once suffer in the polyploid stage, at least before 

 selection can be done to rectify the situation. The sterility barrier is 

 by-passed when a hybridization is included with the doubling; then 

 the degree of fertility generally improves, but not always. The prin- 

 ciple of reduced fertility after polyploidy from the diploid should 

 always be considered by every one starting a new project. Then the 

 changes that might be induced by selection in the later generations 

 can be considered along with the sterility-fertility relations. Granted 

 that fertility levels can be raised by selection, the danger of introduc- 

 ing other changes constantly attends the selection processes. 



The part of the plant to be used for economic production becomes 

 a first consideration, for the root and shoot yields will not be in- 

 fluenced by sterility. Vegetatively propagated plants are a new prob- 

 lem. They need not pass through the reproductive cycle that is so 

 critical to a polyploid at many levels. Perennial plants are favored, 

 and plants that produce propagating shoots like the grasses are im- 

 mediately more favorable than the strictly seed-producing annuals. 



A principle of transfer of characteristics from one species to 

 another has been mentioned frequently in polyploidy work. Among 

 many species the favorable traits are jnomincnt in the wild species. 

 There is at once a desire to introduce this character into the valuable 

 commercial species. A notable case is the mosaic resistance transfer 

 in tobacco. 1" This problem is discussed in greater detail later, but 

 it should be noted that the transfer of such a trait is in effect a prob- 

 lem of polyploidy l)reeding. On a plan in blueprint stage, the idea 

 appears relatively simple, but now it is well known that accomplish- 

 ment is quite difficult. One of the greatest obstacles in transfer is 

 the introduction of undesirable traits along with the desirable ones 

 being sought. 



Combining the good features of two diploid species into the amphi- 

 ploid is another aspect of how hybridization and the doubling of 

 chromosomes offer opportunity for future programs of selection. A 



