520 Mil. II. MAItSIIALL WARD ON 



and cell-division lies at the foundation of the whole, and, therefore, 

 that one cannot expect to recognize a prothallus or a spore &c. 

 from the methods of cell-division. Hence I describe the structure 

 and processes in Butomus in some detail — though it is to the 

 drawings one trusts for the facts ; for they have been made with 

 the greatest care, and are selected from a great number of 

 preparations. 



Embryo-sac of Butomus umbellatus. (Plates XVIL-XIX.) 



* 



An excellent series of figures in Sachs's ' Text-Book ' (3rd edit. 

 p. 489) renders it unnecessary to dwell upon the relations of the 

 ovule to the ovary ; and I may at once describe the structure of 

 the earliest protuberance as it projects perpendicularly from the 

 surface of the placenta into the hollow of the carpel. 



It arises as an emergence from the cells of the carpel-wall, the 

 lining of which is continued, epidermis-like, over it (PL XVII. 

 fig. 1). Prom the first there is a regularity of arrangement of 

 the constituent cells, which becomes better expressed (PI. XVII. 

 fig. 2) in a little later stage, as a central or axial row of cells 

 covered in a dome-like manner by a layer one cell thick, over 

 which in turn the aforesaid epidermal layer* passes. We have, 

 in short, a process of subepidermal cell-division pushing up the 

 epidermal lining of the ovary. The cells for some time divide 

 only by walls perpendicular to the very regular surface of the 

 young ovule ; and each is completely filled with clear or faintly 

 granular protoplasm, the greater part of which forms a large 

 spherical nucleus, with a bright nucleolus in its centre. 



When the axial row has acquired about eight cells, the young 

 ovule-rudiment begins to show signs of asymmetry ; the cells ot 

 the distal half (with reference to origin on the carpel) are larger, 

 and so cause this part to appear somewhat swollen and the whole 

 ovule club-shaped; while certain of the subepidermal cells divide 

 and cause a protrusion on one side. Thus is established the 

 origin of the inner integument (PI. XVII. figs. 3 & 4) ; and this 

 causes the ovule to have its apex directed upwards in the carpel, 

 whence the observer is compelled to cut longitudinal sections of 

 the ovary for the satisfactory preparation of several succeeding 



* There is no advantage in the usage of any special term for the epidermal 

 lining of the carpels. 



