THE EMBttYO-SAC IN ANGIOSPEKM3. 527 



two of the cells cut off by transverse partition of this cell become, 

 by the dissolution of their separating walls, blended to form the 

 embryo-sac — that one of the tetrahedral groups of nuclei is formed 

 by one of these cells, the other by another. Hence, these authors 

 contend, we have a process exactly comparable to what occurs in 

 an anther, where a subepidermal cell cuts off two or more cells 

 at its peripheral end, and becomes the mother cell of pollen- 

 grains, which form in tetrahedral groups of four within each 

 mother cell. Suppose (they argue) two apposed mother cells to 

 fuse, and you have a process substantially similar to what occurs 

 in the formation of the embryo-sac. 



These observations, as far as the embryo-sac itself is concerned, 

 are chiefly due to Vesque ; and the conclusions drawn are that 

 the embryo-sac is not a spore, but a spore- mother cell (or, rather, 

 a fusion of several spore- mother cells), that each of the syner- 

 gidae, the egg-cell, antipodes, and embryo-sac nuclei is morpho- 

 logically a spore. No explanation is offered for the fusion of the 

 embryo-sac nuclei. 



But at the outset we must take exception to Vesque's descrip- 

 tion of the processes, confining ourselves here to a criticism of 

 his remarks and figures on Butomus. 



He admits a difficulty in the case of this plant : though the 

 observation is very easy, the processes do not fit the general 

 theory very readily. He says* a cell of the nucellus enlarges, and 

 terminates inferiorly in a point where it abuts on two subjacent 

 cells ; if the reader compares Vesque's figs. 13 and 14 with Plate 

 XVII. figs. 1-4 of this paper, it will appear that Vesque has over- 

 looked the central row of cells here insisted upon ; a slightly ob- 

 lique section, or a focus too high or too low in the solid cell-mass 

 might give such a figure as his. 



Vesque then figures the first division in this enlarged cell ; and 

 his fig. 15 (with above correction) agrees pretty well with my 

 Plate XVIII. fig. 4 ; but it is not the upper cell which divides 

 again by a parallel wall, as he states, at least in normal 

 cases, but the larger lower one grows and cuts off another cell 

 (cf. Vesque's figs. 16 and 17 with our figures and text). Vesque, 

 however, recognizes a difficulty in deciding this ; and a possibility 

 exists that in our Plate XVIII. fig. 6 some such process has ab- 

 normally occurred. 



■ 



* Ann. des Sc. Nat. 1878. 



LINN. JOURN— BOTANY, VOL. XVII. 2 Q 



