52 
Walter Stiles 
Cellulose walls are generally regarded as completely permeable 
both to water and the vast majority of dissolved substances, the 
chief evidence in support of this opinion being derived from the 
phenomenon of plasmolysis. This is the well-known contraction of 
the protoplast from the cell wall when the cell is immersed in a 
solution of higher osmotic concentration than that of the cell sap. 
Were the cell wall permeable to water but not to the dissolved sub¬ 
stance in the external solution, the cell wall would constitute a semi- 
permeable membrane and so would be the contracting membrane 
to which the protoplast would continue to adhere. 
However, if we accept the view which appears inevitable, having 
regard to the chemical composition of the cell wall, that the latter is, 
at least in part, a colloidal system, there appears every reason to 
suppose that the cell wall may play a part in determining the 
interchange of substances between the cell and its exterior. Thus 
Devaux (1901) showed by a spectroscopic method that the cell 
walls of a number of different species from different groups of the 
plant kingdom possess a strong absorptive capacity so that not 
only the ions of heavy metals such as iron, copper, silver and lead 
are absorbed, but also those of the alkalies and alkaline earths. A 
metal absorbed in the cell wall can be displaced by another. It will 
be recalled that relatively more substance is adsorbed from a weak 
solution than from a strong one, other conditions being equal, and 
Devaux has shown that absorption by the cell wall takes place from 
very dilute solutions. This author (1904) suggested that the absorption 
of substances by the thin walls of the root hairs facilitates the bring¬ 
ing of these substances into contact with the protoplasm of the root 
hair. Since Devaux treated his material with eau de Javclle, weak 
acid and pure water, Cranner points out that he must have removed 
the fatty acids and soaps from the cell wall, so that actually the 
absorptive capacity of the wall must be considerably greater than 
indicated by Devaux’s results, for the lipoid substances he removed 
are known to possess strong adsorptive properties. The assumption 
that the phenomena observed by Devaux are due to adsorption 
(cf. Czapek, 1913; Cranner, 1914) appears not altogether justified by 
the facts; they might be explained as due to chemical action. 
In a series of experiments with parchment membranes and with 
membranes prepared artificially from cell walls, Cranner has shown 
that the amount of water absorbed by a cell wall membrane in unit 
time is influenced by the composition of the solution in which the 
membrane is immersed. The membrane takes up less water from a 
