440 The Ohio Naturalist. [Vol. Ill, No. 8,. 



must be regarded as being bilocular. The related plants, of the 

 AHsmaceae and Naiadaceae, as well as the \'allisneriaceae, are 

 described by Britton and Brown as having two celled anthers. 



Fig. 9 shows the microsporoc\^tes in the pollen sacs. Before the 

 pollen is .shed the grains are tlirce-celled, ha\'ing a large tube 

 nucleus and two smaller male cells (Fig. lo). 



The 3-oung ovule has two integuments ; the inner one is barely 

 beginning to appear when the archesporial cell is organized 

 (Fig. i) The archesporial cell is hypodermal in origin. From 

 this a tapetal cell is cut oiT and this divides into two (Figs. 2 

 and 3). The megasporccyte enlarges and four megaspores are 

 linally produced by two transverse di\-isions, although in some 

 cases some of the walls may be ol)liqt!e (Figs. 4 and 5). The 

 functional megaspore, which is tlie lowest, divides into two cells, 

 forming the two celled embryo-sac in the iistuil way (Fig. 6) ; 

 one cell parses to the upper and the other to the lower end of the 

 sac and by successive divisions the typical eight-celled embryo- 

 sac is f. rmed (Fig. 7). Tl:e polar nticlei are rather large and 

 they conjtigate in about the middle of the sac. A large number 

 of these conjugations was observed, but there was no trace of a 

 triple conjugation of polar nuclei arid a male cell. The syner- 

 giclae lie side b}' side, extending across the upper end of the 

 sac, and beneath them is the oosphere. The antipodals have the 

 same relative position in the lower end that the egg-apparatus 

 has in the upper end. 



At the first division of the definitive nucletis two cells are 

 formed with a trans\'erse wall across tlie sac between them. The 

 upper one of these l)v further division forms a small aiuount of 

 endosperm tissue. The lower one develops into a large vesicular 

 cell with a large nucleus (Fig. 8). This same structure \vith 

 the partition wall was foimd by Schaffner in Saoittaria , by Cook 

 in Castalia and Nymphaca , and by Murbeck in Kiippia. It has 

 also been fottnd in other groups. Contrary to the case in Saoit- 

 taria, no direct division of this nuclues was observed. At this 

 stage the antipodals are in a pocket at the base of the sac, where 

 they persist for a long time. 



The first division of the oospore is transverse. The lower cell 

 elongates and divides transver.^ely also, forming the first three 

 cells of the proembryo (Fig. 11). The upper one of these 

 does not divide but forms a very large suspensor cell as is usual 

 in the Helobiae. The next division is in the middle cell and is 

 also transverse. There is now a tier of four cells. The lowest 

 one of the tier divides longitudinally (Fig. 12), and following 

 this division there is a transverse division in the cell next to the 

 large suspensor cell followed by another longitudinal division in 

 the lowest tier and one in the tier above (Fig. 13). At this 

 stage the pro-embryo is composed of nine cells arranged in five 



