586 SCIENCE PROGRESS 



with increasing temperature so that, other things being equal, 

 the rate of percolation should increase with temperature. 

 This effect has been demonstrated in the laboratory by Bouyou- 

 cos, and in the field by Franklin. The former also showed 

 that, beyond a certain temperature, the rate of percolation 

 reached a maximum and then decreased for soils, while sand 

 showed a continuous increase. He attributes this difference 

 to the presence in the soil of colloidal material, which swells 

 with increase of temperature, thus restricting the diameter of 

 the capillary channels, and leading eventually to a decreased 

 percolation. Possibly the change in structure of the particles 

 which would be brought about mechanically by their expansion 

 in volume when heated would tend to result in closer packing 

 of the grains, and thus decrease the percolation. But this 

 effect was not shown by the sand particles where it might 

 have been expected ; hence the decreased percolation in soils 

 was ascribed to the swelling of the colloidal material. 



Under natural conditions it is probable that the daily 

 fluctuations in temperature induce local oscillations in moisture 

 content in the manner described above. The effect will be 

 to bring to the root-hairs moisture from areas in their immediate 

 neighbourhood, and thus provide a continuous renewal of the 

 soluble substances taken up by the plant. 



(4) The Soil Atmosphere 



Analyses of the gas present in the soil show that it has 

 a slightly greater CO2 content than the atmosphere and corre- 

 spondingly less oxygen. The average volume of COg in the 

 soil air is '25 per cent., and in the atmosphere '03 per cent., and 

 the former shows greater fluctuations than the latter. This 

 comparatively slight difference in composition refers to the 

 " free " soil atmosphere which is withdrawn from the soil by 

 means of a delivery tube inserted to a depth of about 6 inches. 

 In addition to this free gas there is another atmosphere, dis- 

 solved in the moisture and colloidal material of the soil, which 

 can be slowlyremoved by placing the soil in flasks kept evacuated. 

 This dissolved atmosphere contains COg and nitrogen almost 

 exclusively, and practically no oxygen. It is possible, there- 

 fore, for anaerobic and aerobic conditions to exist side by side 

 — a fact which is of great importance in connection with bio- 

 chemical changes in the soil. 



The soil atmosphere is capable of inter-diffusion with the 

 ordinary air, and this process is assisted by the alternate 

 expansion and contraction of the soil gases due to the diurnal 

 temperature wave, as well as by the mechanical displacement 

 effect produced by the percolation of rain into the soil. In 



