Fritsch . — The Moisture Relations of Terrestrial Algae. /. 11 
generally has a slight spiral trend (Fig. 2, c). As drying proceeds the 
striations become somewhat more pronounced. When air-films appear 
around the filaments, the latter exhibit a sudden, rather marked contraction 
in width, and at the same time the striations become very prominent as 
a number of deep furrows and ridges (Fig. 2, e). The transverse contraction 
of the threads amounts usually to one-fifth or one-quarter of the original width, 
though now and again considerably greater (up to as much as five-twelfths) ; 
there is no appreciable longitudinal contraction, and what there is seems to 
set in only after some time. It appears that the delicate longitudinal striae 
seen on the walls of the wet filaments are lines of weakness along which 
the membrane becomes gradually folded during drying, so that the cell-wall 
shrinks round the contracting protoplast and remains closely investing it. 
In dry threads mounted in cedar-wood oil no space is visible between the 
thin wall and the protoplast (Fig. 2,/). According to Imhauser (1889, 
p. 5) the inner parts of the membrane are mucilaginous, a conclusion at 
which I have also arrived. No doubt, therefore, there is a certain con- 
traction of the cell-wall on drying, but there is also a decrease in the 
transverse dimensions of the protoplast to the extent of about one-quarter 
of its previous width, as shown by measurements of the same thread in the 
wet and dry conditions. 
It is noticeable that the septa do not stand out in the dry threads, but 
become more or less arched or zigzagged (Fig. 2, /), a condition frequently 
retained to some extent even in the wet thread, especially if it be of 
relatively considerable width. The characteristic striation of the longi- 
tudinal walls just described has been observed in filaments of this alga 
from diverse terrestrial substrata in the south of England. On the other 
hand, it is not referred to by Imhauser (loc. cit. ), nor have I been able to 
discern it in the abundant Antarctic material of this species which I have 
had the opportunity of examining. It may thus be a special characteristic 
of the extreme terrestrial form. 
The ‘dry threads of the Hormidium stage always show a number of 
prominent knob-like swellings (Fig. 2 ,g), not seen in the wet threads and 
marking points at which the filament has scarcely contracted beyond its 
original dimensions. These are the places at which the characteristic 
‘ concave cells ’ (plano-concave when in pairs) occur which thus resist 
contraction on drying to a much more marked extent than the other cells 
of the threads. The ‘ concave cells * (Fig. 2, d) are always either moribund 
or dead ; at first they are a deep homogeneous green in contrast to the 
lighter green of the ordinary cells, but later on the pigment disappears and 
the cells appear colourless. A chloroplast with pyrenoid is still often 
demonstrable in these cells before they lose their pigment. The concavity 
of the septa adjoining living cells is probably due to the latter pushing 
their way into these cells by virtue of their turgescence (cf. Imhauser, 1889, 
