Thomas . — Notes on Cephaleuros . 785 
radial walls are fully developed before any indication of a transverse septum 
is to be seen (Fig. 4). 
The transverse wall is formed as an annular ingrowth from the lateral 
walls, the lateral and transverse walls being at right angles to each other. 
The formation of this septum must be very rapid, for only rare instances of 
its partial formation are to be seen, although there are numerous examples 
of young radial walls without corresponding transverse septa. 
The most curious feature, however, is that the transverse and longitu- 
dinal septa do not fit. That is, the transverse wall does not join the radial 
wall at its free posterior extremity, but at a point slightly further forward, 
thus leaving a loose end which is a striking feature of the thalli (Figs. 3 
and 4). 
In the early stages a careful examination shows only thin cell-walls, 
colourless or pale yellow in alcohol material, but readily staining with 
haematoxylin, &c. Sometimes protoplasmic contents may be observed, and 
nuclei may be seen in the cells. In older discs the cell- wall consists of 
a double layer, and at a lower focus there are also distinct striations to be 
discerned on the inner wall next the host. 
Rather thick transverse sections of 10 to 13 /x gave the following 
appearances. The thallus cells in section show a large open lumina, 
the wall of the cells staining a pinky violet in haematoxylin. On the lower 
side of the thallus and exactly below the radial walls, a small cell-cavity was 
observed. In oblique sections these may be seen to be hyphae which 
exactly correspond in position with the cell-walls of the thallus, and 
the striations are seen clearly, always on the Algal wall nearest the leaf on 
which the epiphyte grows (Fig. 6). 
Sections cut in various directions support this observation, and the 
hyphae are seen in every case to correspond with the walls of the thallus 
cells, not only with the radial but also with the tangential walls. 
Slides were made by mounting the Alga, after staining, either in 
glycerine jelly or Canada balsam between two thin glass slips, so that either 
surface of the Alga could be examined at will. Examination of these 
slides confirmed what the sections suggest. The lower surface of the Alga 
is covered by hyphae which branch freely, thus covering the lower surface 
with a network, the pattern of which coincides with that made by the cell- 
walls of the Alga. Where these curious hyphae are found in connexion 
with the plant, in a few cases a close connexion between the Algal discs and 
certain pale filaments, which wander in numbers over the leaf surface, 
was observed. The explanation of this curious hyphal structure, so closely 
connected in position with the Algal wall, is not obvious ; no organic 
connexion has been made out between the larger cells of the disc and 
the filaments. 
These may be rhizoids, but the lack of evidence to show a close 
