125 



The embryo sac is not sooner formed, as it seems, 

 than fertilization might occur. Always fertilization is 

 proceeded by the fusion of the polars. Too, the formation 

 of endosperm is somewhat previous to the begin of the 

 growth of the embryo, so the resting oosphère might 

 sometimes be easily overlooked in the rapidly growing 

 mass of endosperm. 



Fig. 12 figures an exception. We feel justified to believe 

 that in this case no fertilization has occured. Nevertheless 

 a considérable growth of the whole sac has taken place 

 as a comparision between fîgs. 11 and 12 will show. The 

 sac is in an indubitable state of degeneration, the scanty 

 amount of strongly vacuolated protoplasm illustrating 

 that fact sufiiciently. The other parts of the ovulum are 

 quite healthy. In spite of the non-occurence of fertilization, 

 the development might, as we hâve found in some rare 

 cases, be continued to embryoless fruit; of course, then, 

 the embryosac sooner or later completely dégénérâtes. 

 Thus Aucuba japonica is added to the growing list of plants, 

 capable of forming parthenocarpic fruits (Tischler 1912). 



In spite of prolonged search we hâve not been able 

 to find in the material at hand the early stages of endos- 

 perm formation. The youngest endosperm seen, contained 

 already sixteen nuclei, each of wich was separated from 

 its neighbours by délicate cell walls. The arrangement of 

 the cells gives the impression that the content of the 

 embryosac has undergone regular division by means of 

 walls since the starting of endosperm formation; but 

 unfortunately we cannot be positiv on that point. As 

 nothing is known about endosperm in Cornaceae, a solution 

 of that question would be rather interesting from many 

 points of view. At the endosperm stage spoken of above, 

 the cells seem to be, at least at the bottom of the 

 embryosac, arranged in two more or less regular rows 

 while upwards their position is more irregular. The lower 



