12 Fritsch . — The Moisture Relations of Terrestrial Algae . /. 
p. 5). The cause of the non-contraction of the 4 concave cells * is still 
obscure ; it may either be due to denser contents or to a firmer wall. It is 
possible that one of their functions is to prevent transverse contraction of 
the threads exceeding a certain limit. At the same time it is probable that 
the cell-membrane at these points of inflation will be in a state of tension 
during drought which may produce a weakening tending to subsequent 
rupture at these places. Certain it is that the frequent fragmentation of 
the threads of this alga not uncommonly takes place at the points where 
such 4 concave cells ’ occur. 
It is quite likely that the 4 concave cells ’ are non-granular ones which 
die away owing to their incapacity to resist drought. It is noticeable that 
such cells are often far more abundant in non-granular than in granular 
threads, although this is not an invariable rule. In material of the same 
alga growing on sandy soil in the Farnham Lane at Haslemere, a more 
rigorous habitat than that at East London College, these cells were very 
plentiful in the threads. 
The behaviour of the threads of Zygnema ericetorum on drying is again 
different to that of the other two terrestrial forms studied. As in the case 
of the Hormidium stage of Prasiola transverse contraction takes place 
suddenly when the air-films appear around the threads, but this is not 
associated with longitudinal folding of the wall along definite lines of 
weakness, although irregular longitudinal folds are often observed. Both 
cell-wall and protoplast shrink, but in the case of the latter this takes place 
to a varying extent. In some threads most or all of the cells show little 
transverse contraction of the protoplast, which remains in close contact with 
the shrunken cell-wall. In others there . is pronounced shrinkage of the 
protoplast, which then becomes separated from the membrane along the 
sides of the cell. In such cells the septa usually remain more or less fully 
stretched, whilst the longitudinal walls cave inwards to some extent, so that 
each cell comes to have somewhat of the shape of an hour-glass. This 
is still more pronouncedly seen in the protoplast, which remains in firm 
contact with the entire width of the septum at either end of the cell, whilst 
the middle part is often very pronouncedly constricted. 
It is worthy of note that the tendency exhibited by these terrestrial 
algae for the membrane on drying to remain in contact with the protoplast 
is also observed in some aquatic forms of this group. In Spirogyra and 
Closterium acerosum the cell-contents usually contract markedly away from 
the membrane on drying, while a narrow species of Oedogonium showed 
changes similar to those just described for Zygnema ericetorum. In the 
case of Cladophora the young green branches generally exhibit marked 
plasmolysis on drying, but in the case of the thick-walled dark-green 
winter stages with dense contents the cell-wall remains closely investing 
the protoplast. This fact is significant in view of the dominance of 
