524 MR. H. MARSHALL WARD ON 



ovule is so simple as here sketched, but that some general law of 

 growth and cell-division ia at the bottom of the question. 



So far as the manner of cell-division is concerned, we cannot 

 argue that the central row differs from any other radial group of 

 similar construction. This row, however, is favourably situated 

 for receiving stores of nutriment to be applied for any special 

 purposes ; and its protected position appears to have been taken 

 advantage of in the formation of an embryo-sac, to which process 

 we now turn. If the ovule of JButomus be compared with that of 

 liosa*, or others where several embryo-sacs appear to occur, the 

 same remarks apply, but apparently more cells at the centre have 

 become utilized. 



As the nucellus grows/and the integuments rise, the somewhat 

 obconical cell in PI. XVII. figs. 5-7 becomes longer and its nu- 

 cleus larger, and then at its front end gives off a cell by the for- 

 mation of a wall across its foremost third or quarter (PI. XVIII. 

 fig. 4). This wall is sharply marked in the centre, but very deliques- 

 cent and swollen at the surfaces bounding the protoplasm. The 

 foremost cell so cut off acquires a nucleus, but does not increase 

 much in size, as does its sister cell below. 



Very shortly the enlarged lower cell repeats the above process, 

 and (PL XVIII. fig. 5) a second cell is cut off in an exactly similar 

 manner. The lower cell again enlarges, and its nucleus becomes 

 bright and round ; and as the cell lengthens, the two sister cells 

 above are seen to become more and more deliquescent as the 

 larger cell in its growth forces them upwards and compresses 

 them against the epidermal cells at the apex. But it may often 

 occur (perhaps always) that the cell first cut off becomes divided 

 by a perpendicular wall into two (PL XVIII. fig. 7), just in the 

 same way that the terminal cells of the other groups divide 

 immediately under the epidermis; in fact, here again it looks 

 as if the greater room for growth immediately under the epidermis 

 were answered to by the freer cell becoming divided. 



In a short time after, however, both the cells cut off by the 

 larger cell are seen to be flattened and rendered a mere refrac- 

 tive cap on this lower cell by its vigorous growth upwards 

 (PL XIX. figs. 2 to 6, &c.) ; and finally the mass disappears alto- 

 gether. At the same time, by a similar process of deliquescence 



* See Strasburger, ' Die Angiospermen u. d. Gymnospermen/ 1879 



