376 Fritsch, Observations on the young plants of Stigeoclonium Kütz. 
sommet et sont plus ou moins inflechis, contournes en divers sens.“ 
In the species, observed by Gay, however, the cells of the rhizoids 
must have preserved their contents better than in 87. variabile, for 
he remarks (loc. eit.): „Les cellules des rhizines conservent leur 
activite physiologique, quelque greles et decolorees qu’elles soient.“ 
The cells of the rhizoids can grow out and form a new vertical 
branch. 
The rhizoids figured by Kützing (53) are of rather a lan 
nature; Tab. 2, fig. IVd shows one of these m St. Zhermale, whilst 
Tab. 8, fig. IT shows a similar one in St. protensum. They are 
relatively very much smaller than any observed by Gay or myself, 
and are further much branched; they rather .resemble some of the 
branched rhizoids found in Oedogonium. 
Wolle (87) also figures short branched rhizoids: in & number 
of species; on Plate C, fig. 1 for Si. fastigiatum; and on Plate CI, 
figs. 8, 9 for St. uniforme var. gracıle. 
Other genera of Chaetophoraceae form rhizoids, very similar to 
those of Stigeoclonium. (cp. Borge 94, Pl. I, fig. 52, for Dra- 
parnaldia.) 
3. The specially modified epiphytic type. 
During the whole of this summer an epiphytic species of 
Stigeoclonium occurred in great abundance on the leaves of a 
specimen of Wallisneria spiralis, contained in the middle tank of the 
water-Äily house at Kew.!) This form ditfered very considerably 
from the usual type, found in Stigeoclonium and I think it will be 
well to describe it immediately and then to discuss its connection 
with the epiphytic species of this genus, already established. 
The first striking point about this species is the extraordinary 
degree of development attained by the base, which constitutes a 
very important part of the whole plant. This basal portion (Tab. XI, 
fig. 14) was usually slichtly concavo-convex in shape, and always 
branched to a great extent. These branches were all firmly 
appressed to the substratum and were frequently so closely packed, 
that their cells almost formed a parenchymatous disc (Tab. XI, figs. 
5 and 11). The vertical branches nearly always originated on the 
convex side (cp. figs.), branching being very rare on the lower side; 
since the vertical branches usually arose from the central portion 
of the delicately-branched base, they radiated outwards in a more 
or less characteristice manner. The cells of the base were ordinarily 
polygonal through mutual pressure, especially when aggregated so 
densely as to form a compact, parenchymatous disc. T'heir contents 
were rich and scarcely vacuolar, a single pyrenoid being always 
very distinet. In the vertical branches on the other hand, in which 
the cells were to a slight extent barrel-shaped, the cell- contents 
were more vacuolar and not infrequently provided with two pyre- 
noids (Tab. XI, fig. 14). 
The mode of development of such a plant is as follows: the 
zoospore on coming to rest assumes an elongated, sometimes slightly 
1) Large numbers of the epiphytie individuals were obtained by scrap- 
ing .off the green filamentous scum, covering.the Vallisneria-leaves. 
