144 Frit sch . — The Morphology and Ecology of an Extreme 
pitted membrane in optical section (Fig. 2, e') and gives a characteristic 
mottled appearance to the protoplast in surface view (Fig. 2, E /r ). If such 
dried filaments are kept in water for a few hours, this distinctive arrange- 
ment of the globules becomes less pronounced. They no longer form such 
a dense continuous layer in all the cells, nor are they quite so closely 
adpressed to the cell-wall, although they still occupy in the main the peri- 
pheral protoplasm (Fig. 2, D). The filaments, subjected to dryness, when 
first placed in water, appear more transparent than after being immersed 
for some time, and this loss in transparency is probably due to the much 
more irregular distribution of the globules in the second case. Not only 
does the arrangement of the globules become less regular in filaments 
immersed in water, but it appears that they also undergo increase in size, 
perhaps accomplished by coalescence of the smaller ones. Lastly, it was 
found that in threads which had been subjected to prolonged drought, the 
globules in many of the cells were partly replaced by one or more highly 
refractive masses of considerable size, again located in the periphery of the 
protoplast (Fig. 2 , F). 
De Bary (’ 58 , p. 10; PL I, Figs. 16, 20) has already described such 
globules in the resting-cells of other species of Zygnema , and refers to them 
as fat-bodies. Their reactions with osmic acid and tincture of alkanna, as 
well as their solubility in chloroform, are quite in accordance with this view 
(cf. West and Starkey, T 5 , p. 198). De Bary does not deal with the varying 
disposition of the fat-globules in dry and moistened filaments, but, seeing 
that these resting-cells are very similar to those of the Hindhead form, it is 
probable that they might show the same phenomena. 
When cells, whether of the dry or wet filaments, are plasmolysed by 
immersion in a strong solution of sodium chloride, the globules in question 
become altogether lost to view. When the plasmolysed cells are allowed 
to recover by placing the threads in water, the globules again become 
visible ; in many of the cells they do not appear to have undergone any 
change, but in some they are not quite as densely arranged as before. 
Plasmolysis does not appear to be effected by anything less than an 8 per 
cent, solution of salt, and, with this strength, but a very slight separation of 
the protoplast from the cell-membrane takes place. In stronger solutions 
(e. g. 10-12 per cent.) some of the cells become more markedly plasmolysed, 
a very obvious contraction of the protoplast taking place, but it requires 
a 15 per cent, solution to produce marked plasmolysis in all the cells. As 
far as I have been able to determine, there is no difference between the dry 
and wet filaments in these respects. Attention may be called to the very 
high osmotic pressures manifested by the cell-sap in this Alga. 
The fact that the peripheral layer of fat-globules remains more or. less 
intact, after recovery from plasmolysis, would indicate a fair degree of 
stability. Whenever there is a shrinkage of the protoplast, such as probably 
