The Ovule and Embryo- Sac in Cassia Tomentosa. 3 



liarity in this particular line of evolution. Cassia itself belongs to 

 the tribe Caesalpiniaceae of the Leguminosae. In Bosa it has been 

 found that the outermost megaspore functions/'' In Eriohotrya the 

 innermost megaspore usually becomes functional, but sometimes the 

 middle or outer one. Occasionally all three begin to develop. i 



Some species of Acacia conform to the normal type, i.e., the inner- 

 most megaspore is functional, but in other species the inner but one 

 of a row of four, or the middle of a row of three, is functional.! 



The development of the embryo-sac, after the stage shown in 

 Fig. v., proceeds rapidly and normally. The four-nuclear sac is 

 shown in Fig. VI. This stage also shows that the functional 

 megaspore is the third from the apex. 



Another peculiarity is found, however, after the next division, i.e., 

 as soon as the eight-nuclear stage is reached. The antipodal nuclei, 

 instead of forming a group at the base of the embryo-sac, he in 

 a row in a tubular prolongation (Figs. VII. a and VII.6). The 

 three lower nuclei are cut off by cell walls, and in considerably 

 later stages are not easily distinguished from the surrounding cells 

 of the nucellus. They do not, however, abort, but appear to 

 function, in common with several cells of the nucellus, as conducting 

 tissue. 



The depleted condition of the cells surrounding the base of the 

 embryo-sac is very noticeable in Figs. VII., VIII., and IX., 

 whereas the antipodal cells have large and deeply staining nuclei, 

 indicating a strong probability that they have absorbed and digested 

 the contents of the surrounding tissue, and will, in turn, keep the 

 egg apparatus supplied with nourishment by this means. At this 

 stage, and until that shown in Fig. X., the micropylar nuclei are 

 exceedingly small, but embedded in rather dense cytoplasm. 



In Fig. VII. a the two middle antipodal nuclei are placed trans- 

 versely, but in Fig. VII. & all four lie in a row. The first case is 

 the same arrangement as Guignard found in Gercis siliquastrum. 

 In each case one of the micropylar nuclei is distinctly separated 

 from the other three, preparatory to its migration towards the centre 

 of the embryo-sac as one of the polar nuclei. In Fig. VIII. the 



* J. Vesque, " D^veloppement du sac embryonaire des Phanerogames Angio- 

 spermes," Ann. Sci. Nat. Bot., 1878 and 1879; Compt. Bend., 88, 1879; Bot. 

 Zeit., 37, 1879. Coulter and Chamberlain, " Morphology of Angiosperms." 



t L. Guignard, " Recherches sur le sac embryonaire des Phanerogames 

 Angiospermes," Ann. Sci. Nat. Bot,, 1882. Coulter and Chamberlain, " Mor- 

 phology of Angiosperms." 



X L. Guignard, " Recherches d'embryog^nie vegetale comparee. I. L^gumi- 

 nenses," Ann. Sci. Nat. Bot., 1881. Coulter and Chamberlain, " Morphology of 

 Angiosperms." 



