112 
Drought 0/ 1870 and 
with mixed mineral manure and ammonia-salts. The third 
3 inches, also, contains more than either; and the fourth more 
than the unmanured, and about as much as the artificially 
manured soil. The quantity continues to diminish to the fifth 
3 inches, and then increases to about the level of the drains. 
To the total depth examined, the dunged soil contained more 
than a quarter of its weight of water, about 3^ per cent, more 
than the unmanured, and about 1 per cent, more than the arti- 
ficially manured soil. 
The soil receiving mineral manure and ammonia-salts also 
retained more water within what may be called the staple than 
immediately below it. It then again increased in percentage of 
moisture, more or less regularly, until within the direct influence 
of the drains. It is to be observed, too, that, whether owing to 
a greater retentive power of the natural clay at that point, or 
more probably to the accumulation, and the action, of the consti- 
tuents of the manures, or of their products of decomposition, 
rendering the clay more hygroscopic, the lower layers of the 
soil of this plot retained considerably more water when saturated 
than did the corresponding layers of either of the other plots. 
The amount of water to the total depth was about 2.V per cent, 
more than in the unmanured soil, but not so much as in the 
dunged soil. 
As might be expected, there are greater irregularities of 
increase or decrease indicated in the percentages of water at 
the different depths, among the results relating to the saturated, 
than among those relating to the dry soils. This may be due in 
part to accidental differences of permeability of the soil, and 
consequently to variation in the freedom of access of the perco- 
lating water, at the different points ; but it is, doubtless, partly 
due to unavoidable error in the collection, weighing, and after- 
manipulation, of soil in so wet a condition. 
Disregarding the irregularities, however, and interpreting the 
obvious direction of increase or decrease of moisture at the dif- 
ferent depths, it is pretty clear that, down to a certain depth from 
the surface — which varied in the different plots according to 
the varying power of retention of the staple and immediately 
subjacent j^layers — the increased percentage of moisture was due 
to the comparatively recent rains. There was then reached the 
layers partially drained since the preceding rains, from which 
point downwards the percentage increased, until again reduced 
by the action of the pipe-drains. 
Further, it is obvious that, by evaporation from the surface, 
and the consequent withdrawal by capillary action of water from 
below upwards on the one hand, and by the gradual descent, 
aided by the natural drainage of the chalk and the artificial 
