322 



rooted in the mud, a new characteristic change occurs in its 

 aerenchyma. At a certain time a layer of cells produced by the 

 phellogen-like meristem will not, hke the preceding ones, develop 

 any columnar extensions, but the walls of the closely packed 

 cells become suberized, and then the whole layer, together with 

 several layers of rounded, also rather closely arranged cells under 

 it, is raised by a new growth of columnar cells of about six tiers, 

 and, finally, another layer, in the same position as the one first 

 mentioned, becomes suberized. Herewith, in the month of 

 October, the growth seems to cease. Further investigation is 

 necessary to learn whether or not, and how, the rootstock will 

 continue to produce aerenchyma the next year. In old root- 

 stocks examined the two layers of corky cells just described were 

 the only ones that could be seen within the aerenchyma. But 

 on older stems, from which the primary bark has partly been 

 thrown ofif, the exposed surface of the aerenchyma is covered 

 with a smooth, dark, blackish-brown pellicle which, when re- 

 moved with the forceps or the razor, proves to be a layer of 

 corky cells. When thin longitudinal or transverse sections are 

 placed in chromic acid, the various Hgnified and cellulose tissues 

 will gradually disappear and these corky layers alone are left 

 The thin membrane will turn over on its flat side, as is usual in 

 such cases, and we can then see that the cork cells are quite 

 closely packed, leaving no intercellular spaces at all. This com- 

 plete separation of the bulk of the aerenchyma from the interior 

 of the stem, and also from the surrounding water, would hardly 

 take place if the plant were dependent upon the aerenchyma for 

 its supply of oxygen. 



In the month of August I found, proceeding from a root- 

 stock, some creeping, runner-like rooting branches, about 8 nun. 

 in diameter. The primary bark was nearly 2 mm. thick and of a 

 spongy structure caused by numerous wide air channels (en. 

 figs- Sf 6). It was separated from the narrow phloem ^one, 

 which was not strengthened by any bast-fibres, by an endoder- 

 mis. Under this endodermis, but extending only over a small 

 portion of the circumference, a meristematic zone of cells was 

 detected resembling those in die floating stem (m. Figs. 5^ ^)- 

 Whether, later in the season, it produced aerenchyma or not, I 



