414 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



of plant quite different from that which the great majorit}^ of the 

 seed would give. Such \arieties are to be regarded as "hetero- 

 zygous" — at least the diploid organization of the somatic cells 

 must be considered as far from "homozygous." The evidence 

 here is therefore in favor of the existence of self-sterility in plants 

 that are heterozygous and whose gametes possess a marked 

 dissimilarity. 



It must be noted, however, that plants that are apparently 

 equally heterozygous may be cither self-fertile or self-sterile. 

 Fifteen plants of the total of 172 plants of the Fi generation re- 

 ported in table 3 were self- fertile. It would seem that there is no 

 evidence of a difference in the degree of heterozygosity between 

 these and their self-sterile sister plants. 



There is also evidence which seems to indicate that self-sterility 

 may exist in plants that are less heterozygous than the Fi hybrids 

 just mentioned. In the inbred strain of the red-leaved Tre^•iso 

 variety of chicory, all plants thus far grown were self-sterile with 

 one exception: the families of this strain have, however, showed 

 wide variations in fasciation, in leaf-shape, and in the amount 

 of anthocyanin developed, and are hence far from a pure race or the 

 characters are showing sporadic variations of the eversporting type. 

 Selection of self-fertile strains derived from the hybrid plants 

 resulted in the segregation of certain F3 progenies that were ^•ery 

 uniform in general appearance and in flower color, and were 

 decidedly uniform in all their characters; in these plants, appar- 

 ently the most homozygous of any of my chicory cultures, self- 

 fertility and self-sterility appeared with quite the same fluctuations 

 as in the Fo generation. We must conclude that complete or 

 partial self-sterility and high self-fertility seem to occur without 

 any direct or immediate relation to those degrees of homozygosity 

 or heterozygosity that have thus far developed in the cross breeding 

 and inbreeding of these plants. It is such fluctuations as these, 

 however, that may be masking the real relations of sterility and 

 fertility to particular types of cell organization or differentiation. 



It is clear that there is no decrease in respect to the self-fertility 

 of strains produced by self-fertilization. The average fertility 

 was quite the same for the F3 as for the F., and in individual 

 records two plants of the F3 gave a percentage fertility of 70 and 56 

 against the highest percentage of ^2 for any Fo plant. The pro- 



