Factors Influencing Plant-Life on Sandy Sea-shores. 43 
conservation of the moisture which has fallen as rain, seem not to 
explain in a satisfactory manner the phenomenon of permanent 
moisture a little below the surface of high dunes, which is noticed 
also in deserts, where no rain has fallen, perhaps for years. 
A more likely interpretation of this fact can be found in 
internal dew formation in the soil. In the layer above the ground 
water, the air which is present between the grains is saturated 
with moisture, and the absolute humidity of this soil-air corresponds 
to the yearly mean temperature of the layer of soil. At a depth of 
60m. this temperature is only about two degrees higher than the 
mean temperature of the surface atmosphere. Now it is known 
that the saturation deficit with the same absolute humidity increases 
with the temperature. Through diffusion the air in the higher 
layers of the sand also tends to be saturated, and any lowering of 
the temperature diminishes the saturation deficit, until a tem¬ 
perature is reached when condensation of water vapour takes 
place. 
While the small gradient in the mean annual temperature 
referred to does not seem to be sufficient to cause a deposition of 
moisture in the interior of the dune, it must be remembered that 
the diurnal and nocturnal temperature variations are considerable 
on an open sand formation, on which the radiation factor is one of 
considerable moment. Such fluctuations in soil temperature are 
periodical and are sufficient to cause periodical condensation of 
water vapour in the soil. While it is inconceivable that the 
apparently permanent moisture in dunes is due to capillarity alone, 
all the foregoing facts are in accord with the theory that the for¬ 
mation of dew in the soil is an additional cause of this moisture. 
The adjustment of soil moisture to the varying conditions of 
the soil is largely regulated by the amount of evaporation. It will be 
noticed on dunes that they are often dry on the surface only a few 
hours after heavy rain. The drier the sand the more slowly perco¬ 
lation takes place, because of the difficulty in overcoming the 
repulsion offered by the dry grain. In fact, during a hot summer-day, 
when the surface is overheated, it will be found that a precipitation 
amounting to 1 cm. has not penetrated deeper than 70 mm. To 
some extent this is due to an evaporation of the water caused by 
the hot grains. But if a low temperature has prevailed for some 
time, even if the sand previously has been quite dry, the water 
percolates to a much greater depth, because in this case the lower 
temperature no doubt had caused condensation of the water vapour 
