216
I.R. Cowan
control system. However, there is a lack of quantitave data to support the
argument. Also, it might be thought surprising if a stomatal response that is
presumably adapted to conserve water were to depend for its realization on
a continuous and uncontrolled loss of water that is not coupled to the
accession of CO2 by the leaf.
With respect to this last comment, and for other reasons, the idea,
espoused in particular by Meidner (1986) and his associates, that the stomatal
response to humidity is due to transpiration from the inner walls of the
guard cell has much to recommend it. Edwards and Meidner (1978) showed
that cuticle is absent on the inner paradermal walls and innermost parts of
the ventral walls of guard cells in Polypodium vulgare, whereas it is thick on
the outer walls of the epidermis and the outer parts of the ventral walls, and
present to varying degrees on the inner walls of subsidiary cells and other
epidermal walls bordering leaf internal air space. It is easy to prove (Cowan
1977) that the inner guard cell walls, provided they are wet, must sustain a
major proportion of the water loss by evaporation within the leaf, simply
due to their proximity to the stomatal pore, even if all the other cell walls
abutting air space within the leaf are also wet. It seems likely that the
observations of Lange et al. with P. vulgare relate to evaporation from
the inner wall of the guard cell, the absence of a humidity response when
the inner walls were in contact with liquid water having less to do with
supply of water to the cell than prevention of water loss from it.
In contrast to the observations with P. vulgare, Nonami et al. (1990)
found that the inner walls of guard cells in Tradescantia virginiana support
cuticle, and argued that transpiration from these surfaces must be very
small, just as they proved it to be from the external cuticle of the leaf.
However, of course, the question is not simply to do with the amounts
of transpiration from the guard cell and subsidiary cell, but the amounts
relative to the hydraulic conductances from epidermal cell to subsidiary cell,
and subsidiary cell to guard cell. Also, it is to be doubted that thickness is a
reliable indicator of the permeability of a cuticle to water. Permeability may
be influenced by the amount and nature of waxes embedded within the
cutin; and it is possible, too, that the permeability of the external cuticle
of leaves is diminished by incipient drying (Sch6nherr 1976; Meidner 1986).
Indeed, Meidner (1976) has demonstrated that the inner walls of the epidermis of Tradescantia virginiana can transpire and be continually replenished with water via the epidermis at rates which are a significant proportion
of rates of transpiration in intact leaves. Water loss from the outer surface of
the epidermis was shown to be much smaller.
There is a weightier objection to the notion that transpiration from the
inner walls of guard cells is responsible for stomatal sensitivity to ambient
humidity. Whatever the proportion the guard cell may sustain of the total
amount of water evaporated within the substomatal cavity and lost through
the stomata, that proportion is unlikely to increase as the stomatal aperture
diminishes (indeed, the reverse is probable). How, then, can closure of the
I.R. Cowan
control system. However, there is a lack of quantitave data to support the
argument. Also, it might be thought surprising if a stomatal response that is
presumably adapted to conserve water were to depend for its realization on
a continuous and uncontrolled loss of water that is not coupled to the
accession of CO2 by the leaf.
With respect to this last comment, and for other reasons, the idea,
espoused in particular by Meidner (1986) and his associates, that the stomatal
response to humidity is due to transpiration from the inner walls of the
guard cell has much to recommend it. Edwards and Meidner (1978) showed
that cuticle is absent on the inner paradermal walls and innermost parts of
the ventral walls of guard cells in Polypodium vulgare, whereas it is thick on
the outer walls of the epidermis and the outer parts of the ventral walls, and
present to varying degrees on the inner walls of subsidiary cells and other
epidermal walls bordering leaf internal air space. It is easy to prove (Cowan
1977) that the inner guard cell walls, provided they are wet, must sustain a
major proportion of the water loss by evaporation within the leaf, simply
due to their proximity to the stomatal pore, even if all the other cell walls
abutting air space within the leaf are also wet. It seems likely that the
observations of Lange et al. with P. vulgare relate to evaporation from
the inner wall of the guard cell, the absence of a humidity response when
the inner walls were in contact with liquid water having less to do with
supply of water to the cell than prevention of water loss from it.
In contrast to the observations with P. vulgare, Nonami et al. (1990)
found that the inner walls of guard cells in Tradescantia virginiana support
cuticle, and argued that transpiration from these surfaces must be very
small, just as they proved it to be from the external cuticle of the leaf.
However, of course, the question is not simply to do with the amounts
of transpiration from the guard cell and subsidiary cell, but the amounts
relative to the hydraulic conductances from epidermal cell to subsidiary cell,
and subsidiary cell to guard cell. Also, it is to be doubted that thickness is a
reliable indicator of the permeability of a cuticle to water. Permeability may
be influenced by the amount and nature of waxes embedded within the
cutin; and it is possible, too, that the permeability of the external cuticle
of leaves is diminished by incipient drying (Sch6nherr 1976; Meidner 1986).
Indeed, Meidner (1976) has demonstrated that the inner walls of the epidermis of Tradescantia virginiana can transpire and be continually replenished with water via the epidermis at rates which are a significant proportion
of rates of transpiration in intact leaves. Water loss from the outer surface of
the epidermis was shown to be much smaller.
There is a weightier objection to the notion that transpiration from the
inner walls of guard cells is responsible for stomatal sensitivity to ambient
humidity. Whatever the proportion the guard cell may sustain of the total
amount of water evaporated within the substomatal cavity and lost through
the stomata, that proportion is unlikely to increase as the stomatal aperture
diminishes (indeed, the reverse is probable). How, then, can closure of the