382 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



In the study of the self-fertility of the Fi, Fo, and F3 generations, 

 comprising a total of 341 plants, controlled self-pollinations were 

 made on 3,140 flower heads, which involved a total of at least 

 55,000 individual fiowers. 



Self- sterility and self-fertility iii the variety "improved red-leaved 

 Treviso.'' — It has already been stated in this paper that as far as 

 tested the plants of eleven cultivated varieties grown in 1913 

 were fully self-sterile. Several plants of the variety red-lea\-ed 

 Treviso grown in that year were fasciated. Cross-fertilized seed 

 was obtained between two of these from which pedigreed progenies 

 have been grown for two generations with the special aim of 

 determining the hereditary behavior of the fasciation. The 49 

 plants thus grown were quite fully tested for self-fertility as shown 

 in the data compiled in table 7. 



The eleven sister plants grown in 1914 were completely self- 

 sterile. The smallest number of heads self-pollinated on any one 

 plant was 8 and the largest 18, but in no case was a single seed set. 

 Cross-pollinations between certain of these plants were successful, 

 giving seed for a crop in 191 5, of which 38 plants were tested for 

 self-fertility. Only one of these plants showed any trace of self- 

 fertility and this plant (no. 21 of series 4, 191 5) was but feebly 

 self-fertile, setting 4, 5, and 8 seeds, judged to contain embryos, 

 in 3 out of 16 heads manipulated. These data indicate that in this 

 variety self-sterility prevails, but that an occasional instance of 

 self-fertility may arise. 



In respect to fasciation, leaf-shape, degree of red pigmentation, 

 and various other characters the plants of each generation showed 

 wide variation. In the Fo generation a few plants were not fasci- 

 ated. They were all quite similar in general vigor, habit of 

 growth, and flower color, but otherwise there were such wide 

 variations that the race must be considered as decidedly impure. 



These plants were derived by inbreeding within a single variety 

 and constitute the only intravarietal breeding that I have con- 

 ducted with chicory unless the one generation of progeny from 

 yl X C is to be thus considered. In both cases, however, the 

 self-sterility is almost absolute and cross-sterility is high. Such 

 evidence raises the question whether cross-breeding between self- 

 sterile plants of strains and varieties that are not of close blood 

 relationship favors the development of compatibility according to 



