344 BOTANICAL GAZETTE [November 



found, at the stage of fertilization, where the archegonium had not been 

 smashed into, either from the top or from the side, by the pollen tube. 



The embryo sac is surrounded by a layer of jacket cells with very 

 large nuclei, and many of the cells are multinucleate. Jager (13) 

 also pictures a very heavy jacket layer of uninucleate cells in Taxus 

 baccata; none are reported in Torreya or Cephalotaxus, although 

 the drawings of the latter by Arnoldi (i) suggest a jacket layer. The 

 archegonia are one to four in number. 



No ventral cell or nucleus could be demonstrated, although the 

 probability is that a ventral nucleus is formed; the chromatin in the 

 egg nucleus {fig. 32b) before fertilization indicates that it is getting 

 ready to divide. Arnoldi (i) says that in Cephalotaxus Forlunei the 

 egg nucleus, shortly before fertilization, cuts off a ventral canal 

 nucleus, which together with a mass of the upper part of the egg 

 destroys the neck cells and passes out of the embryo sac. Jager (13) 

 does not mention nor picture a ventral nucleus in Taxus; Coulter 

 and Land (9) did not find a ventral cell or nucleus in Torreya taxi- 

 folia; Miss Robertson (i) interpreted a spindle in the central cell of 

 Torreya calijornica as the forming of a ventral nucleus. 



Fertilization 



At the time of fertilization the egg becomes rich in food vacuoles 

 in the basal end (figs. 25, 26, 28). The egg nucleus may be situated 

 near the upper end of the egg (figs. 22, 23), or near the basal end 

 (fig. 28). The fusion nucleus (figs. 22, 24) is partially surrounded 

 by the finely granular cytoplasmic sheath of the male nucleus. This 

 cytoplasmic sheath has been observed in Taxodium by Coker (8), 

 in Torreya taxijolia by Coulter and Land (9) , and in T. calijornica 

 by Miss Robertson (19). The non-functional male cell (figs. 22, 23), 

 which has begun to disorganize, shows the cytoplasmic sheath very 

 distinctly. 



Embryo 



The first division of the oospore could not be obtained. Free 

 nuclear division was observed to the eight-nucleate stage. As this 

 was the oldest stage available, it is impossible to say whether more 

 free nuclei are formed or not before the formation of walls. Arnoldi 

 (1) figures ten free nuclei in one section of Cephalotaxus Fortunei, 



