stout: pollinations in cichorium intybus 413 



tion divisions undoubtedly gave in each plant many kinds of 

 pollen with respect to the actual qualitative values of the germ 

 plasm (irrespective of whether or not there may be purity of 

 segregation). Fifty-three of these plants were fully self-sterile; 

 all kinds of pollen produced by each plant were ineffective in self- 

 fertilization, though in respect to segregation the pollen of these 

 plants must, it would seem, be as markedly different as that 

 from the five sister plants that were, however, self-fertile in some 

 degree. In the F2 and F3 generations, self- fertile and self-sterile 

 plants appeared quite independently of the individual hereditary 

 complex, so far as it could be judged by visible characters. The 

 class of white- flowered segregates of the F3, which were very 

 uniform in appearance for each line, included both self-sterile 

 and self-fertile plants. Likewise, the members of F3 progenies 

 grown from blue-flowered parents were self-sterile or self-fertile 

 without respect to flower color or other characters. These results 

 are quite the same as those obtained by Correns, East, and Morgan, 

 which indicates that complete self-sterility may operate in indi- 

 viduals that are highly "heterozygous," and that in such cases 

 the processes prohibiting self-fertility are operating on the pollen 

 tubes or gametes from pollen of that plant irrespective of their 

 idioplasmic differences, so far as the latter can be judged by visible 

 characters. In cross-fertility or -sterility quite similar behavior is 

 seen. It is on such evidence that the views of Jost and Morgan, 

 that the processes are purely somatic and intercellular in the 

 individual itself, are based and if self-sterility were universal in 

 hermaphrodites this would be a very natural conclusion. 



The marked persistence of self-sterility in varieties propagated 

 vegetatively, already considered in the cases of the apple, plum, 

 pear, cherry, and blueberry, is suggestive that the conditions 

 giving self-sterility are perpetuated by bud propagation without 

 any marked change due to growth under somewhat changed con- 

 ditions. Carefully conducted experiments with such plants are 

 greatly to be desired to determine whether wide variations can 

 occur in the plants of the same clone Irrespective of, or in associa- 

 tion with, evidences of somatic variation in nuclear organization. 

 If we may judge by the seed progeny of the cultivated varieties 

 just mentioned, the evidence is quite clear that they are vegetative 

 types that propagate fairly true vegetatively and maintain a type 



