252 W. HOFMEISTEB ON THE DEVELOPMENT OF ZOSTERA. 



the germinal vesicle into a small lenticular, lower cell, and a 

 larger, expanded, upper cell (fig. 25). The latter contains no 

 nucleus ; a thin layer of granular mucilage coats the inside of 

 its wall, and its cavity is filled with watery fluid. This upper 

 cell is not further developed during the subsequent completion 

 of the seed. 



The little lower cell, on the contrary, at once enters upon a 

 very rapid multiplication. It swells up into a flattened globular 

 form, and divides by a longitudinal wall ; both the newly- formed 

 hemispherical cells then immediately divide into two " daughter- 

 cells,'^ having the form of quarters of a sphere, by septa stand- 

 ing at right angles to those just before produced. These four 

 cells, constituting the rudiment of the embryo, are each divided 

 by a horizontal cross septum (fig. 26). By continued halving 

 of its cells, principally in the longitudinal direction, the flat- 

 tened globular mass of cells (fig. 26) soon becomes spherical 

 (figs. 27, 28) ; and finally, while constantly increasing in size, 

 ovate and compressed laterally (fig. 29). The large spherical 

 cell which supports it is only loosely attached to the inside of 

 the wall of the micropyle end of the embryo-sac. I more 

 than once saw the swollen cell, and the cellular mass sprung 

 from it, slip from the micropyle end of the embryo-sac half- 

 way towards the other extremity, without perceptible external 

 cause*. 



By the time the rudiment of the embryo has acquired the 

 globular form, a single apical cell becomes distinctly visible 

 (fig. 28), in the organ which in previous stages of development 

 undoubtedly possessed four. This transformation may be aptly 

 explained by the supposition that one of the four cells originally 

 forming the apex divides into one inner and two outer cells; 

 either by a longitudinal septum forming an angle of 45° with 

 each of the two side walls of the cell, after which a radial lon- 

 gitudinal wall appears in the outer of the newly-formed cells ; 

 or by a longitudinal wall parallel to one of the lateral surfaces, 

 the formation of which is immediately followed by that of an- 

 other at right angles to it. If we further suppose that, in either 

 case, the other three of the previously apical cells become di- 

 vided by radial longitudinal septa, during the formation of one 



♦ See the note to fig. 26 in the explanation of the figures. 



