140 Frit sc h . — The Morphology and 'Ecology of an Extreme 
than by a piece of filter-paper of equivalent size. These phenomena are 
due to the fact that the greater part of the thick wall is mucilaginous, so 
that the dry filament swells up very considerably on wetting. Thus, the 
average width of the dry threads is 16 fi, of the wet threads 24 // ; the 
average thickness of the longitudinal walls of dry filaments (examined in 
spirit) is 2*3 j u, of wet filaments 5 //. 
It does not appear that the inner layer of the wall is at all appreciably 
mucilaginous, since it retains its outline and normal appearance, even in 
dry threads. On the other hand, the intermediate layer must be very 
largely gelatinous in character in the Hindhead form, as it is very much 
shrunken and often quite thin in the dry filament. In all probability the 
mucilaginous character increases from the interior towards the outside, the 
surface layer being very strongly mucilaginous (cf. above). 
The advantage of this mucilaginous wall to the Hindhead Alga is 
obvious. It not only involves a very rapid recovery of the cells, when 
supplies of moisture become available, but, since the gelatinous walls will 
only part with their water by slow degrees, the full effects of a drought will 
only be felt by the Alga some time after dry weather sets in. It is indeed 
astonishing how long the desiccated filaments still retain their soft flexible 
character. Moreover, at certain periods of the year, although exposed to 
intense insolation in the daytime, such a form will be able to make use 
to the full of the dew deposited at night, so that even during periods of 
drought it may be able to hold its own to some extent. In short, we see in 
this character a pronounced adaptation to the conditions of the habitat. 
In the material of the ordinary terrestrial form at my disposal, the inter- 
mediate layer was generally much thinner and did not appear to be 
appreciably mucilaginous. . 
A considerable number of aquatic Zygnemaceae form gelatinous sheaths 
about their filaments, but these are sharply marked off from the cell-wall, 
and there is reason to believe that they are composed of mucilage excreted 
from the cell-contents (Klebs, ’86, p. 368) and not, as in the Hindhead 
form, a product of the membrane. West and Starkey (T 5 , p. 194) record 
the presence of an occasional more or less distinct gelatinous sheath in 
Z. ericetorum , but I have never met with anything of the kind in the 
Hindhead form. Thick and stratified mucilaginous membranes are, on the 
other hand, frequently found in the resting-cells (akinetes) formed, as a result 
of desiccation, by many species of Zygnema (. Z . leiospermum, Z . pectinatum) 
(de Bary, ’ 58 , pp. 9, 10), so that it might be said of the Hindhead Alga 
that it existed permanently in the akinete condition. 
The development of the septum between two daughter-cells is initiated 
by an annular invagination of the innermost layer of the wall, about mid- 
way between the two chloroplasts or the two halves of the chloroplast 
(Fig. 1, b). It often has the appearance as though this were due to 
