Rosendahl: embryology of symplocarpus foetidus. 3 



sliows that during a normal winter season there is practically no 

 growth taking place between these two dates. In the latter figure 

 the archesporial cell has enlarged somewhat and the nucleus is pre- 

 paring for division. According to the observations of Mr. Gow 

 the plants behave quite differently in New Jersey, where no trace 

 of the inner integument is to be noted before the latter part of 

 February, and the archesporium does not become recognizable be- 

 fore the first or second week of March. 



Upon resuming growth in the spring the successive stages in the 

 development of the embryo-sac follow each other rapidly. The sin- 

 gle archesporial cell (Fig. 4-5, PI. I and Fig. 30-31 PI. Ill) becomes 

 the spore-mother cell, which by two successive divisions gives rise to 

 four megaspores, and not three as recorded in the above mentioned 

 observations (Fig. 6-10, PI. I). The position of the four mega- 

 spores with relation to each other is not always the same in different 

 ovules. In most cases the two innermost ones lie in the long axis 

 of the ovule, while the two outermost lie transversely (Fig. 8-9, 

 PI. I). This arrangement of the four megaspores is very likely 

 what has given rise to the erroneous statement that only three spores 

 are present, for if the section happens to strike through the ovule in 

 such a way that the two outermost spores lie over each other it is 

 easy to overlook one of them (Fig. 9, PI. I). More rarely all four 

 spores are arranged in a straight row (Fig. 10, PI. I). At this 

 stage in the development of the megasporangium the inner integu- 

 ment has pushed forward but does not cover the end of the nucellus. 

 The outer integument is just beginning to appear in the form of a 

 slight swelling about on a level with the base of the nucellus. 



It seems probable that all four spores germinate. However this 

 stage in the development has not been observed in a sufficiently 

 large number of embryo-sacs to fully substantiate the statement. 

 The nuclei of three spores have been seen to be in a state of division 

 (Fig. II, PI. I). The fourth, one of the outer or micropylar ones, 

 still remaining undivided, although showing signs of preparing for 

 division. (In the figure only two dividing nuclei are shown, the 

 third, being in a deeper plane and obscured by the undivided one, is 

 in such a position that it could not be drawn). 



The successive nuclear divisions in the development of the embry- 

 o-sac from the spore evidently follow each other with great rapidity, 

 for from a very large number of slides examined, the two- and 



