THE EMBRYO 271 



angiosperms is that of Moringa oleifera, in which Rutgers (1923) 

 claimed that wall formation commenced only after the 16-nucleate 

 stage. Puri (1941) has shown, however, that the free nuclear em- 

 bryo of Rutgers was really a micropylar accumulation of endosperm 

 nuclei, while the real embryo, which is formed quite normally, was 

 entirely overlooked by him. 3 



In the earlier stages of development there are no fundamental 

 differences between the embryos of the dicotyledons and those of 

 the monocotyledons. However, since the mature embryos are so 

 markedly different in the two groups, they will be treated separately 

 in the following account. 



DICOTYLEDONS 



Except in a very few species, which will be considered later, the 

 first division of the zygote is almost always followed by the laying 

 down of a transverse wall. Of the two cells thus formed, the one 

 which lies towards the interior of the embryo sac is called the 

 terminal cell and the other the basal cell. In the next stage the 

 terminal cell may divide transversely or longitudinally. The basal 

 cell usually undergoes a transverse division, but in some plants it 

 remains undivided and becomes hypertrophied to form a large 

 vesicular structure. 



The French embryologist Soueges, who is the chief authority on 

 the development of the embryo in angiosperms, considers the mode 

 of origin of the four-celled proembryo and the contribution made by 

 each of these cells to the body regions of the mature embryo as the 

 most important aids in a classification of the embryonal types. Fol- 

 lowing him, Schnarf (1929) and Johansen (1945) have recognized 

 five principal types of embryos among the dicotyledons. These 

 may be distinguished from one another as follows: 

 I. The terminal cell of the two-celled proembryo divides by a 

 longitudinal wall — 



(i) The basal cell plays only a minor part or none in the sub- 

 sequent development of the embryo Crucifer type 4 



3 The free nuclear divisions of the zygote reported in Ficus (Tischler, 1913) and 

 Ruta (Cappaletti, 1929) are in the nature of abnormalities which need not be 

 considered here. 



4 Johansen (1945) prefers the term "Onagrad type" as the embryo of the Ona- 

 graceae is simpler and more typical than that of Capsella (Cruciferae) , usually 

 described in textbooks. 



