536 MB, II. MARSHALL WARD OS 





The importance of the cell-wall (in PL XXV. fig. 8) passin 

 across the sac must not be overrated ; it is a rare case, and is 

 only insisted on (together with the instances quoted above) be- 

 cause the general character of the successive divisions across the 

 embryo-sac mother cell suggests that they belong to one series 

 that the mother cell becomes divided, in fact, into four superposed 

 cells. But the fact that in most cases no thin wall is recogni- 

 zable does not prevent our accepting the probability of such a 

 wall existing ; it is in a very diffluent state, but is still there, and 

 by its absorption of fluid appears like a vacuole between the 

 nuclei in the sac. 



In my next figure (PL XXV. fig. 10) the embryo-sac has elon- 

 gated and filled up the whole space left by the integument as it 

 closed up to form the micropyle ; the nucleus at the micropyle 

 end has formed (by alternate division into 2x2) an " egg-appa- 

 ratus" — two " synergidse " closely packed into the apex, and an 

 " egg-cell" suspended from their bases; while the fourth nucleus 

 is seen travelling down the protoplasm-plates to join its fellow 

 given off from below. In the base of the sac are three " anti- 

 podal cells " (cf the account of JButomus). 



The completion of the sac &c. is shown by figs. 11 and 12 to be 

 quite e& in Butomus : the two nuclei fuse (fig. 11); and a large 

 bright "nucleus of the embryo-sac " (fig. 12) results. The cells 

 around the base of the sac (in fig. 12 especially), and forming a 

 tube in which the " antipodal cells " lie, are the lower set of epi- 

 dermal cells of the nucellus (cf. PL XXV. figs. 9-12), which 

 appear to become quite vigorous again. 



As in the Composite, the cells of the integument around the 

 sac become arranged with reference to it, and give a characteristic 

 appearance to the mature ovule. 



Emlryo-sac tifc. of Verbaseum phlomoides. (Plate XXV.) 



One more example of a nucellus reduced to the " axial row " of 

 cells and an epidermis is furnished by such an ovule as is figured 

 at PL XXV. figs. 2 & 3. The young ovule (fig. 2) has a large 

 cell terminating the axial series; and the integument grows 

 forwards with it, till, in fig. 3, we have a row of four cells pro- 

 duced by the division of the large granular cell in the apex of 

 fig. 2. The middle wall of the three is the oldest ; having divided 

 it into two, each cell then divides again. "Whether the embryo- 

 sac arises from one or more cells in fig* 3, 1 cannot yet state; 



