The Co-Efficient of Humidity. 
Water-Content. 
i 29 
In terms of wet soil. 
In terms of air-dry soil. 
10 % 
25 
50 
70 
80 
85 
12-5% 
33-3 
100 
233-3 
400 
566-6 
All values quoted in the paper are expressed in terms of the air- 
dry soil. 
III. —Variability of the Water-Content. 
So far the water-content has been assumed to be the only 
measure of the soil-moisture, as in fact it has hitherto been regarded. 
But it has already been implied that it proves to be a very imperfect 
index of the soil conditions, not because it varies, for the soil 
moisture is itself subject to variation, but because it varies irregularly 
and independently of the humidity of the soil. Broadly, the water- 
content will by its magnitude suffice to distinguish habitats if the 
soil is uniform to some depth, but that condition is rarely satisfied in 
natural habitats and can never be postulated. Still, there is no 
difficulty in differentiating a heather moor and a cotton-grass moor 
by the mean water-content of the peat, but in the majority of cases 
the evidence it affords is inconclusive or contradictory. Even on a 
heather moor, peat containing only 70 or 80% of water can be shown 
to be in reality wetter than peat containing 100 or 130% of water. 
But the real difficulty in the use of the water-content is met 
with as soon as the section sampled is not of the same nature at 
successive depths. Where soils are shallow, as where peat overlies 
a coarse sand that rapidly passes into rock, or where a woodland 
humus covers a soil almost free from humus, the active rootlets may 
occupy both layers indifferently, though, as shown by Woodhead 1 
and others, in certain cases these several layers are occupied by 
distinct members of the association. It was soon discovered that 
in analyses of such shallow soils the water-content diminished so 
rapidly at the boundary between the layers that the values were in 
no wise comparable, in spite of the fact that both related to the 
same plant. A few illustrations selected from and typical of the 
earlier stage of this investigation will make the point clear. 
' “ Ecology of Woodland Plants in the Neighbourhood of Huddersfield.” 
Journ. Linn. Soc. Bot., Vol. XXXVII., p. 333. 
