Permeability 57 
With regard to the position of the semi-permeable layer very 
diverse opinions have been expressed. From the experiments with 
Hordeum vulgare, var. ccerulescens, and by following the penetration 
of sodium chloride by silver nitrate, A. J. Brown thought it possible 
that the walls of cells derived from the nucellus form the semi- 
permeable membrane in barley grains. Schroeder in the case of 
wheat suggested the possibility of a localised entrance of water and 
dissolved substances into the grain, or possibly decreasing per¬ 
meability of the coverings from the basal to the apical end of the 
grain. Reichard (1909) correlated the presence of a tannin layer in 
the barley grain with semi-permeability. Shull (1913), from his 
investigations on a number of species, particularly Xanthium glabra- 
turn, concluded that cellulose walls may form semi-permeable mem¬ 
branes, and stated that “the possible semipermeable character of 
cellulose membranes cannot be overlooked in future investigations 
dealing with the entrance of salts into plant tissues.” 
Shull disposed of Reichard’s view that tannin determines the 
semi-permeable character of a layer in the seed coats of Gramineae 
by extracting the tannin from grain with sodium hydroxide without 
affecting the semi-permeable properties, while Collins (1918) was 
unable to confirm the presence of a tannin layer in barley grains. 
The question appears to have been solved for the case of the 
barley grain by Collins who, in a careful morphological and physio¬ 
logical investigation, showed that the barley grain is completely sur¬ 
rounded by a strongly cutinised membrane except in the region of 
the micropyle and the chalaza. Only a very small proportion of the 
water absorbed by the grain can pass the cutinised layers, while 
dissolved substances are also kept back. The seat of rapid entry of 
water is at the micropylar region of the grain, and here must be the 
seat of semi-permeability. Entry of water through the chalaza 
appears much slower. The cells in the micropylar region (the so- 
called embryonic appendage) through which water must pass, appear 
possibly to be modified in the direction of pectic, mucilaginous, or 
gum compounds. The conclusion is not definite, and so cannot be 
used as evidence either for or against Shull’s opinion of the possible 
wide distribution of semi-permeable cellulose membranes in plants. 
From this review of the properties of cell walls with regard to 
permeability it is clear that we may find walls with very different 
degrees of permeability both to water and to various dissolved sub¬ 
stances. 
(To be continued) 
