22 



the embryo do not begin until ;i \;\\-<yc number of endosperm 

 nuclei have been produced. The first segmentation is 

 transverse, and the resulting cells become separated by a 

 marked basal wall (Fig. 27). The lower or hypobasal cell 

 lies next to the micropyle, and simply grows into a large 

 suspensor cell. The epibasal or apical cell gives rise to 

 all the members of the embryo. The order of its divisions 

 appears to follow the course described by Cannon for the 

 development of the embryj in A-rt-na fntua} as given b\' 

 Coulter and Chamberlain.-' " In this species the cotyledon 

 and stem-tip are both derived from the apical cell, the 

 entire root-tip (including root-cap) from the adjacent cell, 

 and the coleorhiza from the third cell, the suspensor con- 

 sisting of only the primary basal cell." Whilst these 

 features may be illustrated in wheat by reference to an 

 embryo plate of six cells, it must not be concluded that the 

 course has been actually traced and demonstrated, the 

 gaps in my material haxing prexented this from being 

 accomplished (Fig. 2.S). 



,sr is the large suspensor cell, the separating line i- i 

 is the original basal wall ; of the complex mass of five cells 

 above that line, cii represents the cotyledon mitial, .v the 

 stem, r the root, whilst co probably gives rise to the 

 coleorhiza or root sheath ; 2 2 is the first transverse wall 

 of the embryo cell, 3 3 its first longitudinal one. 



This order of segmentation does not appear to be 

 nlways followed ; other embryos may have longitudinal 

 walls succeeding each other at right angles giving rise 

 to a solid mass of cells. 



Figure 29, taken from an embryo of 14th July, 1900, 

 shows not only the different parts much more clearly 

 marked off, but the primary tissues, dermatogen, periblem 

 and pleromc, being differentiated. 



