432 University of California Publications in Botany [Vol. 5 



this connection we may consider the three following possibilities in 

 the ease of the species under consideration. In the first place we may 

 think of a ripe ovary containing at least two hundred ovules, each one 

 properly matured and capable of the production of a viable seed fol- 

 lowing self-fertilization. After such fertilization and under normal 

 conditions the seed capsule is retained on the plant up to and follow- 

 ing the shedding of the ripe seed. On the other hand, another ripe 

 ovary of the same plant in which no ovules are fertilized falls from 

 the plant along with the other flower parts and a portion of the pedicel 

 a few days after such maturity. Finally, in the case of a third ripened 

 ovary on the same plant the fertilization of two or three ovules is 

 sufficient to inhibit flower-fall. The fact that in experiments 14 

 and 17 in the above table but two and five ovules respectively could 

 have been fertilized^ is evidence in this connection. Additional evi- 

 dence is found in the way in which the fruits of the F^ Tahaciim- 

 sylvestris hybrids remain upon the plant when only a very few seeds 

 are finally matured in them. In this case cytological examination of 

 ripe ovaries showed that but a very small number of the six or seven 

 hundred embryo-sacs are normal and capable of fertilization. When 

 the viable 'pollen of either of the parent species is placed on the 

 stigmas, the majority of the flowers thus pollinated remain attached 

 to the plant and a few seeds are fully or partially matured in each 

 ovary. We have here a case in which normally all the flowers on the 

 plant fall as a result of lack of pollination with viable pollen and in 

 which it is actually impossible for more than a very few ovules to be 

 fertilized, yet when these few fertilizations are accomplished abscission 

 does not take place. 



Throughout the above discussion we have emphasized fertilization 

 as contrasted with final formation of viable seed as the determining 

 factor in the abscission of the flower. This has been done advisedly 

 with the following facts in mind. In the F^ Tabacum-sylvestris 

 hybrids ripe seed capsules retained on the plant often contain nothing 

 but a few empty seed cases. In other words it would seem that the 

 fertilization of all or a number of the normally matured embryo-sacs 

 provided the stimulus necessary to inhibit the activation of the separa- 

 tion layer, the question as to whether or not any of these fertilizations 

 resulted in the formation of viable seeds being non-essential. Further, 

 the results of experiments 14 and 17 in the above table indicate that 



3 The results in general indicate that branched pollen tubes with the neces- 

 sary sexual elements do not occur in the style and thus that only one ovule is 

 fertilized by a single pollen grain. 



