316 BOTANICAL GAZETTE [APRIL 
If the cellulose membranes which are adjacent to the air spaces 
of the leaf and those bathed by the outer air were structurally 
_ similar to the filter paper used in the test just described, the above 
figures might be applied to the moist foliar surfaces. But we may 
be sure that filter paper possesses a much less compact structure 
than any cell membrane in the plant; hence, the reduction in the 
evaporation rate brought about by partial drying out of the moist 
cell walls should be much more marked in the case of the latter 
than is manifest in our experiment. It seems quite probable that 
a reduction of 50 per cent in the water content of exposed cellulose 
walls should produce an equal or much greater reduction in the 
evaporation rate therefrom, and in the case of cutinized epidermis 
(which holds but little moisture when saturated, and from which 
cuticular transpiration takes place), the effect of partial drying out 
should be still greater. 
It appears, therefore, quite within the limits of possibility, that 
the phenomenon of incipient drying of exposed membranes (inter- 
nal and external) may be adequate to cause the non-stomatal 
hindrance to transpiration here considered. It should make no 
difference whether the excessive rate of water loss from foliage 
leaves be brought about by high evaporating power of the air, by 
absorption of sunshine, by continually rising temperature, or 
(merely relatively) by a decrease in the possible rate of water 
supply (as by the drying of the soil or by the removal or injury of 
the basal part of the plant), the effect must be the same in every 
case, namely, a marked fall in the rate of relative transpiration. 
If such a process of drying out in leaf tissue were continued, 
water would eventually be extracted from the protoplasmic mem- 
branes, which would in turn remove moisture from the vacuoles, 
and (since we suppose that the rate of supply from the vascular 
elements to be inadequate) the turgor pressure would decrease. 
Finally, all internal pressure would be removed from the cell walls, 
which would thus cease to be under strain, and all turgidity would 
thus be destroyed; the wilting point would be reached. Con- 
tinued still further, the process would result in actual plasmolysis 
of the cells and finally in death, after which desiccation would 
rapidly take place. The latter portion of this supposition seems 
