114 INTRODUCTION TO EMBRYOLOGY OF ANGIOSPERMS 



tributed and very similar to one another in appearance. Some 

 other workers also obtained similar results, and noted that the 

 mature embryo sacs occasionally showed more than three antip- 

 odals. D'Amato (1940a), Ekdahl (1941), and Walker (1950), work- 

 ing on several species of Ulmus, have clarified the position by show- 

 ing that as a rule four divisions intervene between the megaspore 

 mother cell stage and the organization of the embryo sac, and not 

 three. The coenomegaspore shows a 1 + 3 arrangement of the 

 megaspore nuclei, each of which undergoes two further divisions, 

 resulting in the formation of 4 nuclei at the micropylar end and 12 

 at the chalazal end. Frequently, however, some of the chalazal 

 nuclei fail to undergo the fourth division, resulting in a total of 14, 

 12, or only 10 nuclei, of which 4 are at the micropylar end and the 

 rest at the chalazal. Several of the latter degenerate soon after 

 their formation, so that there is a further decrease in the number of 

 nuclei, and eventually only two to four antipodal cells may be differ- 

 entiated. Also, in certain cases the 4 megaspore nuclei divide only 

 once, so as to give rise to an 8-nucleate embryo sac of the Adoxa 

 type. 



The embryo sac of Chrysanthemum parthenium presents a range 

 of variation which seems to indicate that there are several races of 

 this plant which behave somewhat differently from one another, 

 although possibly the differences are related to environmental con- 

 ditions. According to Palm (1915), who gave the first detailed 

 account of the embryo sac of this species, each of the 4 megaspore 

 nuclei divides twice. The 16 nuclei arising in this way organize 

 to form a three-celled egg apparatus, two polar nuclei, and eight 

 antipodal cells of which the basal cell is four-nucleate. 



Fagerlind (1941) studied two specimens of the same species. In 

 specimen 1 the 4 megaspore nuclei were observed to take up the 

 most variable positions, and frequently the 3 basal nuclei were 

 seen to lie in close contact (Fig. 735). With the subsequent elon- 

 gation of the sac the nuclei became separated from one another by 

 vacuoles, the micropylar nucleus being larger than the rest (Fig. 

 73C). All the nuclei now divided simultaneously, resulting in 8 

 nuclei, of which the 2 basal were the smallest and soon began to 

 degenerate (Fig. 73D-F). When the next division (Fig. 73G) was 

 over, there were 14 nuclei in the sac, of which 3 organized into an 

 egg apparatus, 2 functioned as polar nuclei, and the rest formed 



