We can only consider briefly the root-air relation, since little work has been done on the soil atmosphere. We therefore do not know the rate of movement of air in the soil, or, for that matter, its composition. It is probable that there is a large per cent, of carbon dioxide where there are a relatively large number of roots of plants, bat as to the diffusion of oxygen from the air into the soil or the diffusion of carbon dioxide from the soil into the air, little appears to be known. Preliminary tests show that there may be more movement of the air in the soil than might at first be supposed, and that varying, even if small, atmospheric pressure may directly affect air movements in the soil. For examples, if a tube 50 cm. long and 2 cm. in diameter be filled with soil composed of sifted sand and adobe—one part of the former to two parts of the latter—it will be found that a water pressure of only 1 cm., or less, will be required to force a continuous stream of air through it. The pressure given is for soil saturated with water. When air-dry, there is almost no resistance. In soils of this composition, therefore, it is probable that ordinary variation in atmospheric pressure isto induce in it rapid air movements. Preliminary experiments, in which a stream consisting of 20 c.c. of air a minute was passed through the soil where the roots were placed, indicated by the great vigor of the plant, and the relatively extensive root development, that that amount of air was beneficial to development and forwarded growth. Variations in temperature with depth of soil, variations in water content of the soil, are both additional potent factors in modifying the rate of movement of the soil air.
While it is not known in an exact way how the atmosphere of the soil effects the position or certain other features of the root-systems of plants, it seems probable that in certain cases, at least, the effect is pronounced. For example, as has been shown above, the root-systems of the cacti without exception are placed near the surface of the ground. The roots grow in a soil horizon which is not the most moist, but, on the other hand, which although moistened first is also the first to give up its water, and it very likely is the optimum air content of the soil at a critical period which determines the superficial placing of the roots. It is a well-known fact that many bulbous plants require well-drained soil, which is probably only another way of saying that they thrive best in soils having good aeration. Two or three experiments may be cited which may be taken to substantiate the conclusions just stated. For example, there grows in the vicinity of Tucson a cylindro-opuntia (Opuntia arbuscula?) in which the root-system is fleshy, the roots having much the appearance of slender sweet potatoes. It was supposed at first that the fleshy roots of these species was a specific character, which, indeed, may be true, and therefore obligate. Some doubts, however, have been thrown on this conclusion from observations on another