92 Fertility in Cichorium intjbiis 



the widely different results in this case being obtained from almost 

 identical ancestral records. After three generations of self- fertile parents 

 with a record of 4, 13, and 32, one rather large series of 29 plants gave 

 31 °/„ of plants self-fertile, but with ranges only to 26%! ^^id the very 

 low average of 8°/^. A sister series with ancestral record of 4, 22, and 

 31 °/„ gave 50 7,, of jjlants self-fertile, of a range extending to 02°/^ and 

 an average of 25 °/„. Such irregular and sporadic results seem character- 

 istic so far as my data go. 



However, on the whole, this family was one of relatively low ancestral 

 fertility. Only three of the members have shown a fertility above 26 ^/„, 

 and the average of 0165 is for all offspring lower than that of any other 

 family. 



The record of ancestry for the family {E22 x A)-10- is one of high 

 fertility. The fertility of the first self-fertile parent was 51°/^. The 

 23 plants of the first generation, 16 of the second, and 82 of the third 

 all had self- fertile parents with no fertility lower than 25°/^, yet of 

 these the percentage of self- fertile plants was 41, which was almost 

 identical with the record for the whole family. The most highly 

 self-fertile series yet obtained in regard to the percentage of self-fertile 

 plants was in this family (see {E22 x A)-10-8-15-), but there were also 

 two series that were almost completely self-sterile. 



In the family {A x E3:i)-9-, the plants selected for parents have 

 been, as a rule, of medium fertility, although, of course, these have been 

 above the average. The range of fertility is considerably higher in the 

 family, and the average fertility of self-fertile plants is also higher than 

 in any other family. 



Certain aspects of the results are perhaps clear. In every family, 

 as a rule, the total fertility (proportion of plants .self-fertile, range of 

 fertilities, average percentage of fertility) is increased over that whicli 

 evidently occurs in the progeny of self-sterile parents ; at least the pro- 

 poi'tion of self-fertile plants is much greater. This increase is, as a 

 rule, marked in the first generation of offspring grown from self-fertile 

 parentage. Continued selection for parentage high in individual self- 

 fertility does not steadily and continuously increase the fertilities of the 

 progeny either as to individual or to average fertility. The marked 

 tendency to the development of self-incompatibility has not been 

 eliminated. 



An inspection of the data presented will show that in several series 

 the numbers of plants that are self-fertile are relatively high. For 

 example, in series having 83, 78, and 77 '/^ of the plants self-fertile it 



