240 
Walter Stiles 
permeability is proportional to the rate of linear shrinkage, in the 
second to the rate of linear shrinkage multiplied by the square of 
the length. Hence if comparisons are only made, as in the case of 
Miss Delf's experiments, with cylinders in the same stage of shrink¬ 
age, l 2 will be the same in all cases compared with one another and 
the permeability will similarly be proportional to the rate of linear 
shrinkage. 
The values obtained by Miss Delf may then legitimately be 
regarded as measures of the permeability of the cell membranes 
(cell wall + protoplasm) to water. Miss Delf speaks of them as 
measures of the permeability of the protoplasm, but as the effects 
of cell wall and protoplasm are not separated, they are better re¬ 
garded as measures of the permeability of the cell. It will be noted 
that errors arising from exosmosis of dissolved substances and intake 
of salt are allowed for in the experimental procedure, and need not 
therefore be further considered. 
Determinations of the Influence of Temperature on the Rate 
of Intake of Water by the Change in Weight Method 
Storage Tissues. The influence' of temperature on the intake of 
water by storage tissues, potato tuber and carrot root, has been 
investigated by Stiles and Jorgensen (1917 b). In the method em¬ 
ployed by them, the course of absorption of water at different 
temperatures is followed by immersing the tissue of known weight 
in water and weighing the tissue at intervals. The change in weight 
may be fairly attributed to gain or loss of water only, as the weight 
of dissolved substance which diffuses out from the tissue during the 
period of the experiment is negligible in comparison with the weight 
of water involved. 
In order to eliminate errors due to the variability of tissue in 
different samples, a variability which may be very large, and to 
ensure equality of surface and weight of tissue in all comparative 
experiments, the following procedure was adopted. Cylinders of 
tissue were cut from the tubers or roots by means of a cork- 
borer i-8 cms. in diameter. These cylinders were then cut into 
disks of a uniform thickness of 0-2 cm. by means of a hand micro¬ 
tome. For any series of experiments performed for comparative 
results all the disks required were cut at once, well mixed together 
and weighed in sets of 20. In this way samples of the material, each 
sample consisting of 20 disks, were obtained having approximately 
