6fa°. REPORT—-1899. 
supposing every one of the stomatic openings of this leaf could be filled up with a 
solution of caustic alkali, the absorbent power of the leaf as a whole would only 
be #4 of what it actually is when assimilating. 
These are some of the consequences which flow from an acceptance of the 
hypothesis of stomatic exchange, and it appeared to be impossible to accept that 
hypothesis unreservedly without some collateral evidence that these comparatively 
nigh velocities of diffusion are physically possible when dealing with such low 
gradients of tension as must necessarily exist when the highest amount of carbon 
dioxide does not exceed ‘03 per cent. 
The well-known general law expressing the rate of the spontaneous inter- 
mixture of two gases when there is no intervening septum, was, as every one 
knows, established by Graham, and the more elaborate investigations of Loschmidt 
many years later served to confirm the general accuracy of this law, and to show 
that, within very narrow limits, the diffusion constant varies in different gases 
inversely as the square roots of their densities. 
But a mere knowledge of the diffusion constants of air and carbon dioxide does 
not, as far as I can see, materially assist us in the particular case we have under 
consideration. In order to gain some idea of what is actually possible in the way 
of stomatic diffusion in an assimilating leaf, we must know something of the actual 
rate at which atmospheric carbon dioxide can be made to pass into a small chamber 
containing air at the outside tension, but in which the carbon dioxide is kept down 
almost to the vanishing point by some rapid process of absorption; and we must 
also determine the influence of varying the size of the aperture through which the 
diffusion takes place. 
Our attempts to answer these questions experimentally have led us into a long 
investigation, which promises to be of wider interest than we had first imagined. 
I only propose to give on this occasion a general account of the results so 
far as they affect the physical question of the intake of carbon dioxide into the ~ 
plant. 
When a shallow vessel containing a solution of caustic alkali is completely’ 
covered, the air above the liquid is very speedily deprived of the whole of its carbon 
dioxide. If we now imagine a hole to be made in the cover of the vessel, carbon 
dioxide will enter the air-space by free diffusion, and its amount can be very accu- 
rately determined by subsequent titration in the manner I have previously referred 
to. The time occupied by the experiment and the dimensions of the aperture being 
known, we can express the results in actual amounts of carbon dioxide passing 
through unit area of aperture in unit of time; or, since the tension of that gas in 
the outer air is known, we can express the average rate of the carbon dioxide 
molecules across the aperture in terms of actual measurement, say centimetres per 
minute, 
We have made a very large number of experiments of this kind, using, in the 
first instance, dishes of about 9 cm. in diameter, and varying the size of the holes 
in the cover, the air-space over the absorbent liquid being always the same. 
The accompanying curve, fig. 1, illustrates the effect which a gradually 
decreasing orifice has on the rate of diffusion of atmospheric carbon dioxide under 
these conditions. The diameters of the orifice in millimetres are given on the 
abscissa line, and the rates of diffusion through equal areas of the apertures are 
taken as ordinates, the rate of absorption in the open dish under similar conditions 
being taken as unity. 
lt will be seen that in the first instance a gradual reduction of the diameter of 
the opening is accompanied by a very regular increase in the rate of passage of the 
carbon dioxide until a diameter of about 50 mm. is reached; that is to say, up to a 
point at which about two-thirds of the area of the dish is covered. A further 
progressive diminution in the size of the aperture makes comparatively little 
difference in the diffusion rate until we reach about 20 mm., beyond which the 
om again begins to rise, increasing rapidly in steepness as the apertures become 
smaller, 
The experiments with open dishes are too crude for a study of the influence of 
very small apertures, so for this part of our work we constructed a special form of 
