Correspondence. 24t 
The co-efficient may be deduced if we wish, from the figures in the 
first and third lines; but apart from that the soils all look different and 
the middle one decidedly wetter than both the others; also the loss 
at roo°C in the first case seems disproportionately large and suggests 
incomplete drying at 15°C. 
(B) In terms of the Air-dry Soil. 
Soil No. 172 1731 173 ii 
per cent. per cent. per cent. 
Water-Content .. 24.8 125.00 29.1 
Loss as 100° C. 5.92 6.34 2.10 
humus ..; ate 25.28 25-37 5-33 
Residue .. be 68.80 68.25 92.55 
I00.00 99.99 99.98 
We are now on much surer ground for we are looking at the soils apart 
from the water they hold—the quantity of which may vary from day to day. 
The first two are as alike as it is possible for two soils to be and both rich 
in humus, whereas the third is deficient in it; further, the apparent 
excessive loss of water at 100°C. has vanished. So the analysis stated in 
terms of the wet soil was both deceptive and failed to bring out the true 
character of the soils. 
But we have not yet arrived at a true estimate of the wetness of the 
soils, for the middle one looks much wetter than the others. Applying 
water content 
the figures in either table to find the ratio | —,_ we get :— 
umus content 
172 Tyo 173 li 
co-efficient of humidity 0°98 4°9 574 
We conclude that the habitat of the hairgrass (No. 172) is a very dry 
one, so much so that only a drought resisting plant could survive the 
conditions. But even more interesting is the way in which the lower 
section of No. 173 responds to the test “and reveals itself as a thoroughly 
wet soil quite in line with the upper portion, and in strong contrast to 
No. 172, though there is no disparity in their water-content. The cor- 
rection already mentioned would reduce the value of 5.4 somewhat and 
bring it even nearer to 4.9. But for a discussion of this I must refer those 
who are interested to my paper on the co-efficient of Humidity. It is 
evident that this co-efficient furnishes a new standard for judging the state 
of humidity of a soil and as such it is likely to be of value in Ecology. 
There remains only one little point that requires elucidation. A single 
oe 
Minerz 
way of expressing the humus-content of the peat. it. is not a matter of 
any importance, but I have found it convenient in dealing with long series 
of soil analyses to have one number in sight by which to classify them. 
Humus 
Mineral 
0.06 respectively, and these numbers are the proportions of humus to one 
part of mineral residue in each. If the number is greater than 1.0 the soil 
may be classed as a fairly pure peat or its equivalent; if on the other 
hand it is less than o.1 the soil is decidedly deficient in humus ; if less than 
0.01, it is of the nature of dune sand. 
line in the ‘ Abstract’ is given to the ratio as a convenient 
Thus in the soils already quoted the ratios are 0.37, 0.37, and 
W. B. Crump. 
HALIFAX, 
May 18th, 1913. 
1913 June 1. 
