stout: pollinations in cichorium intybus 357 



of 207 such combinations 48 were fertile in some degree. There 

 were wide variations, however, in the individual behavior of the 

 various group combinations, as can be illustrated by the following 

 summary compiled from tables 8 and 9 : 



In general, as in the crosses with parents, the cases of cross- 

 fertility exceeded expectations; out of 299 cases that should have 

 been cross-sterile, 113 combinations were fertile. In the tables 

 Correns includes data on self-pollinations of 13 plants and of these 

 it appears that three (2 d, 2 e, and 2 h) were partly self-fertile; in 

 five of the plants reported self-sterile only two flowers were tested, 

 in two only three, in one plant four, and in one plant six. No 

 doubt a larger number of self-pollinations would reveal more plants 

 self-fertile. It also appears from the data of table 8 A that at least 

 one plant included in the tables (i c) was not even tested for self- 

 fertility. There is no data given regarding the self-sterility of the 

 other 46 plants of the Fi generation. 



Correns' results are important in establishing the fact that cross- 

 incompatibility may exist between closely related plants which 

 also exhibit self-incompatibility. His contention that both self- 

 and cross-sterility are due to the same kind of interactions is 

 suggestive. That this incompatibility depends upon and exists 

 between line stuffs whose presence depends on an "anlage" which 

 is inherited and which follows the Mendelian law of segregation is, 

 however, not supported by his evidence. It is plain from Correns' 

 data that such a simple explanation is not adequate, even if we 

 believe that incompatibility is largely dependent upon chemical 

 compounds and definite processes. 



These results show very clearly that there are wide variations 

 in the degree to which cross-sterility exists among different plants, 

 that conditions giving sterility are not identical in all plants and 

 that when complete cross-sterility exists it may be quite indepen- 

 dent of germ-cell differences resulting from reduction divisions. 



