I08 THE SCOTTISH BOTANICAL REVIEW 



what is presumably the primary endosperm nucleus, but the 

 further history of this was not followed. 



Figs. 30 and 31 show details from an embryo-sac of A. 

 simplex somewhat more advanced than that figured in 24. 

 In this case the egg apparatus was near the chalazal end of 

 the sac and the endosperm filled the whole cavity. In this 

 sac the first formation of endosperm had evidently taken 

 place at the chalazal end and proceeded toward the upper 

 part of the sac which was filled with a few large cells. The 

 group of smaller endosperm cells at the chalazal end looks 

 somewhat like antipodals, but these cells are probably not to 

 be considered as essentially diff"erent from the other endo- 

 sperm tissue. In one of the larger cells there was a group of 

 large nuclei of which two are shown in the figure, and these 

 look as if they were preparing to fuse into one. A similar 

 group of smaller nuclei was seen in another cell nearer the 

 apex (fig. 31). Actual fusion of nuclei was repeatedly 

 observed in somewhat older stages, and it is evident that 

 this fusion of endosperm nuclei is a common phenomenon of 

 no special significance (see figs. 35, 39, 40). 



As the embryo-sac enlarges subsequent to the first divisions 

 of the embryo, the walls of the larger endosperm cells become 

 firmer and the sac is filled with a solid mass of large cells with 

 clearly defined walls. There can usually be made out on the 

 concave side of the sac a somewhat more compact mass of 

 tissue which probably represents the primary endosperm, but 

 it is not apparent that this is essentially different in its nature 

 from the later-formed and larger-celled tissue. 



The early history of the embryo-sac in A. inodestuvi differs 

 somewhat from A. simplex. The young sac, as we have seen 

 (figs. 28, 29), completely destroys the apex of the nucellus 

 and reaches a large size before the first nuclear division occurs. 



The egg apparatus in this species in most cases was placed 

 at the chalazal end, this perhaps being associated with the 

 fact that the apex of the nucellus is destroyed and the upper 

 part of the embryo-sac is entirely exposed. Fig. 35 is a 

 section of the embryo-sac with the chalazal egg apparatus {0) 

 and two free nuclei {e, n). These latter probably divide 

 again, as in A. simplex^ before the first walls appear in the 

 endosperm. 



