A. B. 8T0UT 101 



Ten flower heads self-j)olIinated on a completely self-sterile plant will 

 set no seed, while ten heads on the same plant pollinated on the same 

 day with pollen from a highly cross-compatible plant will set abundant 

 seed. The fi-uits are rather small achenes having no endosperm, and are 

 practically composed only of the embryo : provided the pollination is 

 comjjatible, they develop equally well throughout the season (Data, see 

 1916, Tables XV and XVI). 



8. Self-Gompntibility and self-incovvpatihUitij appear independently 

 of any combination of germ ])lasm elements in so far as these can be 

 judged by the expression of characters. Each operates alike between 

 gametes that are similar or those that are dissimilar in resjjcct to 

 hereditary units of genetic analysis. Plants widely different in such 

 qualities as cohjur of flowers, type of branching, shape of leaves, etc. 

 are either self-fertile or self-sterile, and plants of a sister series quite 

 similar in all respects are either self-fertile or self-sterile. When an jP, 

 plant of hybrid origin is self-fertile in any degree the evidence indicates 

 that any of the sex cells may function in any recombination ; on the 

 other hand in self-sterile sister jDlants whose sex cells must, it would 

 seem, be of much the same diversity none are compatible. Also all the 

 sex cells of an F^ plant which must have much the same germ-plasm 

 constitution may fiiil to function together, while those of a sister plant 

 may be highly functional. Two self-sterile plants, sisters of an F^ cross 

 or sisters of any generation, may be cross-fertile or cross-sterile quite 

 indiscriminately. 



9. The development of either self- compatibility or self-incompati- 

 bility occurs in both cross-bred and inbred races, the latter often being 

 highly constant races for vegetative characters. Both self-fertile and 

 self-sterile plants occurred among sister plants that were F^ hybrids of 

 rather wide crosses (Stout, 191G, 1917); they also appeared among 

 inbred strains derived by crossing self-sterile parents for as many as 

 three generations (1917), and they occuri-ed, as here reported, among 

 the progeny of self-fertile plants, even after three generations of self- 

 fertile parentage. The positive evidence at hand, however, makes it 

 clear that self-compatibilities do not decrease in self-fertilized lines 

 of descent which are so uniform that they constitute decidedly pure 

 races. 



10. The results obtained in the cultures of chicory make it clear 

 that self-incompatibility and self-compatibility are here not to be described 

 as dominant and recessive characters, or paired allelomorphs, and that 

 there is no simple Mendelian formula that fits the results. The evidence 



