CHROMOSOME NUMBERS AND THEIR ALTERATION 209 



the same pole, (see i)age 106). ('hiasinata liberate the chromatids from 

 this restriction in regions far from the kinetochores; hence in such regions 

 they are (hstributerl at random. On the basis of such random distribu- 

 tion of eight chromatids in these regions, the ratios for genes located 

 there would be the following. After selfing: AAA A, 1:0; AAAa, 783 : 1 ; 

 AAaa, 21:1; Aaaa, 2.48 : 1 ; aaaa, 0:1. After a testcross the correspond- 

 ing ratios would be 1:0, 27:1, 3.7:1, 13:15, and 0:1. In a number of 

 researches expected ratios have been found to be rather closely approxi- 

 mated (Fig. 155). 



The breeding behavior of autopolyploids differs markedly from that of 

 diploids because of the relative infrequency with which recessive char- 

 acters reappear in succeeding generations. Thus the F^ generation is 

 more uniform than in diploids.* Another fact of genetical and practical 

 interest is that tetraploid lines in many cases are kept distinct from the 

 diploids by cross incompatibility and, in case fertilization is accomplished, 

 by the failure of the resulting triploid lines to compete successfully with 

 the diploids and tetraploids. 



Triploidy.^ — Triploid plants almost always arise from the union of a 

 haploid gamete with a diploid one, the latter having been produced regu- 

 larly by a tetraploid plant or after a failure of haplosis in a diploid plant. 

 Such plants, like tetraploids, show good vegetative growth and are fre- 

 quently somewhat larger than the diploids. Although the sexual fertility 

 of some triploids is low because of the irregular meiosis where three 

 genomes are present, others, e.g., maize and iris, show good fertility'. 

 Progenies derived from triploid plants tend to have a low sur^'ival "\-alue 

 because of their aneuploid chromosome numbers, and this greatly restricts 

 the value of triploidy in the development of new types. As a general rule 

 triploid plants are unsuccessful in nature unless they have some form of 

 asexual reproduction upon which they can rely; hence they are rarely 

 found established among sexually reproducing species in the field. 

 Numerous highly valued plants of the orchard and garden, e.g., certain 

 varieties of apples, tulips, iris, and hyacinths, are triploid, but they are 

 normally propagated ])y vegetative methods. 



Higher Degrees of Polyploidy. — Species with 6, 8, and 10 or more 

 genomes are found in nature and show a high degree of fertility. That 

 their establishment probal)ly was accomplished gradually and may ha\e 

 involved some hybridization is suggested by experimentally produced 

 autopolyploids with such numbers of genomes. In maize, to select an 

 example from a considerable number of known cases, induced octoploid 

 plants are far less vigorous than the tetraploids and are completely sterile. 

 Again, when doubling is induced in commercial varieties of potatoes, 

 which are tetraploids, the resulting octoploid plants and tubers are inferior 

 to the tetraploids; whereas, when related diploid species are doubled, the 



