﻿1 903] EMBR YOGENY OF ZAMIA 1 89 



archegonium may be amitotic; after the number of free nuclei 

 has become quite large, the vacuole is formed by the disorgani- 

 zation of the cytoplasm and free nuclei of the central portion of 

 the ^gg. After the vacuole has become estabHshed there 



remains at the base of the ^gg a rather dense mass of cytoplasm 



containing numerous free nuclei, while the cytoplasm in other 

 parts of the tgg^ is reduced to a thin peripheral layer with com- 

 paratively few nuclei. At this period free simultaneous nuclear 

 division is resumed in the region at the base of the ^g%, which 

 is apparently quite sharply marked off (Ikeno, 8,/^. ^p), so 

 that there are two rather definite periods of free nuclear division. 

 In Zamia [figs, j,^, /o) there is only one period of free nuclear 

 division, and no vacuole is formed during these early stages or 

 even after the suspensor has begun to elongate, the free nuclei 

 of the upper portion being scattered throughout the cytoplasm. 

 Even in stages more advanced than that shown in jig, 10 a 

 vacuole has not yet formed, although the cytoplasm is becoming 

 ery scanty in the region which in Cycas is occupied by the 



large vacuole. 



G 



{fig 



fig 



and free nuclei, becomes empty. 



In the formation of cell walls also Zamia differs from Cycas. 

 In Cycas circmalis walls are formed throuarhout the entire region 



occupied by nuclei, while in Zamia the formation of walls is 



restricted to rather a limited area at the base of the e.gg. The 

 earlier walls {Jig. y) are formed as in the endosperm of those 

 angiosperms in which a period of free nuclear division precedes 

 the formation of cell walls, excepting that in Zamia we have 

 never found more than one nucleus in a cell, as sometimes 

 occurs in angiosperms. At first the cells are all of the same 

 size, but very soon those at the base begin to stain more deeply, 

 and on account of more rapid division become smaller than those 

 above them [Jig. 8). The differentiation soon becomes quite 

 marked, and the elongating cells of the suspensor can be dis- 



tinguished easih 



{fig. 9)- In 



fig^ 



that just above the definitely walled cells there is a region where 



