584 THE AMERICAN NATURALIST [Vol. XLII 



Soon this mass of cells bends a trifle downwards, and the 

 anlage of the sporangium appears as a swelling in the 

 distal axil formed by the juncture of the sporophyll with 

 the receptacle (PI. I, Figs. 8 and 10). The end of the 

 sporophyll now begins to bend upwards and the ovule 

 downwards. About this time the archesporial cell ap- 

 pears as a single large hypodermal cell (PI. I, Fig. 13). 

 When the ovule has bent downwards about 90° and the 

 seed coats have begun to appear at its sides as slight 

 elevations, the archesporium divides rapidly to form four 

 megaspores, arranged approximately in a longitudinal 

 row. The first two, however, often lie more or less side 

 by side (PI. I, Fig. 11). In a few cases I found only 

 three megaspores. This may have been because of tardi- 

 ness in the division of one of the cells. Meanwhile a 

 groove has been forming in the upper surface of the 

 carpel, and as the carpel continues to bend upwards the' 

 ovule is entirely enclosed by the sides and bottom of the 

 groove. By the time the megaspores are formed, the 

 upper convex surface of the ovule has fused with that 

 portion of the pistil next to it. The ovule now hangs 

 loosely downwards in the cavity thus formed. The 

 nucellus continues its circular movement and by the time 

 it has bent 180 the megaspore next to the chalaza, the 

 functional one, has enlarged, while the others have begun 

 to degenerate. When the ovule has described an arc of 

 225° the functional megaspore has made its first division. 

 Between the two nuclei a large vacuole, which continues 

 to enlarge with the growing sac, is formed (PI. I, Fig. 12). 

 The cells at the sides of the sac crowd in between the 

 sac and the epidermis. At this stage the seed coats are 

 prominent and have three or four layers of cells. In the 

 next stage the nucellus has advanced about 90° farther in 

 its cycle and each nucleus of the sac has divided. With 

 the increase in size of the sac, the cells round about it 

 begin to degenerate. The seed coats are now as long as 

 the nucellus (PI. II, Fig. 14). While the nucellus is 

 passing through the next 45° of arc all the nuclei of the 



