Tue CARBON DIOXIDE OF THE SorIL AIR 327 
activities of the soil, and this accumulation of carbon dioxide will, by the 
mass action law, finally result in a retardation of further production of 
the gas. On the other hand, they maintain that by drawing a current 
of air continuously over the surface of the soil, conditions more nearly 
similar to those obtaining in the field will result. This may be true for 
experiments conducted in a quiet room; but in the greenhouse, where there 
is a circulation of air, there is ample opportunity for diffusion to take 
place from the soil, especially where, as in these experiments, one of the 
lower cocks of the soil can was always left open, so that a sample taken 
at any particular time should be truly representative of the carbon dioxide 
actually present under normal conditions. 
It has been pointed out by Leather (1915) that usually only about 
25 per cent of the carbon dioxide in the soil is in the gaseous state, 
the remainder being dissolved in water. It is reasonable to suppose that, 
once the soil water is saturated with this gas, any further production of 
carbon dioxide will tend to increase the content in the soil air. Considering 
these facts, then, it will be seen that the method used in these tests will 
not give, and was not intended to give, absolute amounts of carbon dioxide; 
but it nevertheless should yield reliable relative values. 
On April 2, 1917, the soil, which is a heavy clay loam rich in silt and 
having a lime requirement of about 3000 pounds to the acre (Veitch), 
was brought up to 30 per cent moisture content on the oven-dry basis. 
Four of the cans were seeded. to White Russian oats. A half-inch layer 
of quartz sand was then spread over the surface of the soil in the eight cans. 
From April 12 to September 28 the samples were taken twice a week. 
From September 29 the sampling was done approximately once in two 
weeks until February 7, 1918, after which date the samples were again 
taken twice a week. The second crop of oats was planted on January 9. 
Some fifty seeds were usually sown, and the plants were thinned out in 
the course of two weeks to fifteen in each can. In the season of 1917, one 
plant became infected with smut, and on June 13 this plant was removed, 
together with one plant from each of the other cans. To maintain the 
moisture content of the cropped cans at 30 per cent (oven-dry basis) 
frequent waterings were necessary, especially at the time of most vigorous 
growth. At that period the cropped cans were irrigated once a day. 
The amount of water added was recorded in order to see whether or not 
there was any relationship between the transpiration and the carbon- 
