674 , REPORT—1899, 
If, instead of taking the air from a height of three or four feet from the ground, 
we examine the stratum only one or two centimetres above the surface of a soil 
free from vegetation, we find, as might be expected, a very large increase in the 
amount of carbon dioxide, which may exceed, under these circumstances, 12 or 13 
parts-per 10,000 of air. Such a.soil, in fact, gives off by diffusion into the sur- 
rounding air an amount of carbon dioxide which is comparable to that evolved 
by a respiring leaf, that is to say about 50 c.c. per square metre per hour. This 
is probably a factor which has to be taken into account in considering the assimi- 
lative power of vegetation of very low growing habit, but in all other cases we 
may assume with safety that aérial plants have to take in their carbon dioxide 
from air in which its tension does not exceed ;5%5, of an atmosphere. 
The actual intake of carbon dioxide is determined by enclosing the entire leaf 
in specially constructed air-tight glazed cases, through which a sufficiently rapid 
air stream is passed. These cases are so arranged that the leaf can be enclosed 
whilst still attached to a plant which is growing out in the open under perfectly 
natural conditions, and some of them are sufficiently large to take the entire leaf 
of a sunflower. 
The carbon dioxide content of the air is determined both before and after its 
passage through the apparatus, and since the amount of air passed is known we 
have all the data requisite for, determining the actual amount retained by the 
leaf. 
An experiment generally lasts from five to six hours, and the carbon dioxide 
fixed in this time may amount to 150 c.c. or more, the actual error of determina- 
tion being very small indeed. 
For purposes of comparison the volumes are reduced to the actual number of 
cubic centimetres of the gas absorbed by a square metre of leaf in one hour, which 
of course necessitates an exact determination of the area of the leaf. This is most 
conveniently effected by printing the leaf on sensitised paper, and tracing round 
its. outline with a planimeter set to read off square centimetres—a far more 
accurate and expeditious plan than that of cutting out a facsimile of the leaf from 
paper of a known weight per unit of area. 
If it is desired to estimate the assimilative power of a leaf in an atmosphere 
artificially enriched with carbon dioxide, the air stream before entering the leaf 
case is passed through a small tower containing fragments of marble, over which 
there drops a very slow stream of dilute acid, whose rate of flow is so propor- 
tioned to the air stream as to give about the desired enrichment with carbon 
dioxide. The stream of air is then divided, one part going on directly to the leaf 
case, whilst the other passes through a separate absorption apparatus and meter 
for the accurate determination of its carbon dioxide content. 
In order to show the kind of results obtained in this manner I will give one or 
two examples. 
A leaf oi the sunflower, having an area of 617°5 sq. c.m., was enclosed in its 
case whilst still attached to the plant, and was exposed to the strong diffuse light 
of a clouded sky for five and a half hours, air being passed over it at the rate of 
nearly 150 litres per hour. The content of the air in carbon dioxide as it entered 
the apparatus was 2°80 parts per 10,000, and this was reduced to 1:74 parts per 
10,000 during its passage over the leaf. This corresponds to a total absorption of 
139°95 c.c. of carbon dioxide, or to an intake of 412 c.c. per square metre per 
hour. If we assume that the average composition of the carbohydrates formed is 
that of a C,H,,0, sugar, the above amount of carbon dioxide corresponds to the 
formation of 0°55 gram of carbohydrate per square metre per hour. But we must 
bear in mind that the average tension of the carbon dioxide in the leaf case was 
only equal to 1-98 parts per 10,000—that is, only about seven-tenths of its tension 
in the normal air. A correction has therefore to be made if we wish to know how 
much the leaf would have taken in, under similar conditions of insolation, if it had 
been bathed with a current of air of sufficient rapidity to practically keep the 
amount of carbon dioxide constant at its normal amount of 2°8 per 10,000. We 
shall see later on that, well within the limits of this experiment, the intake is 
proportional to the tension, so that applying this correction we may conclude that 
