20 
INVESTIGATIONS ON ROTH A.MSTED SOILS. 
Table 5. — Broadbalk wheat soils sampled in October, 1865; October, 18S1, and 
October, 1893 — Percentages of stones, fine dry soil, stubble and roots, etc., and 
moisture in soils as sampled. 
SUMMARY, EXCLUDING DUNGED PLATS. 
Depth. 
Number 
of plats. 
Total 
stones, a 
Fine dry 
soil. 
Stubble, 
roots, 
etc. 
Mois- 
ture. 
Samples collected October 2-9, 1865: 
First 9 inches 
Per cent. 
Per cent. 
Per cent. 
Per cent. 
in 
16. 16 
73. 78 
0.05 
9.01 
Second 9 inches - 
10 
11.03 
72.68 

16.29 
Third 9 inches. 
10 
6.34 
75. 57 
18.09 
©ampies coiiecteu. v^clouoi iu-io, ±001. 
19 
14. 77 
68.89 
.03 
16.31 
19 
8.98 
71.54 
.01 
19.47 
Third 9 inches 
19 
5.24 
72.77 
.01 
21.98 
Samples collected October 13-21, 1893: 
19 
13.02 
70.92 
.01 
16.05 
19 
9. 66 
71.05 
.01 
19.28 
19 
5.18 
73.33 
73.38 
21.49 
12 
6.05 
20.57 
Fifth 9 inches 
12 
4.76 
74.26 
20.98 
Sixth 9 inches - 
12 
12 
12 
4 
4 
6. 15 
7. 82 
6.12 
2.74 
5.03 
73.20 
72.26 
73.99 
78.83 
77.31 
20.65 
19.92 
19.89 
18.43 
17.66 
Seventh 9 inches 
Eighth 9 inches 
Ninth 9 inches 
Tenth 9 inches 
a Stones retained by i-inch sieve. The tine dry soil includes stones passing through i-inch 
sieve. 
The short summary given in Table 5, especially in the 1893 division, 
shows at a glance the average composition of all the plats, excluding 
the dunged plats, at the various depths. There is a general though 
not uniform decline in the proportion of stones as we go deeper, 
and, of course, a corresponding increase in fine soil, together with 
a general increase downward in the moisture retained by the soil, 
though in the last depths there is a decrease. The variations are, 
for a considerable depth, due, no doubt, mainly to the causes of 
weather, vegetation, etc., already alluded to, but in the lower depths 
the varying proximity of the underlying chalk must to some extent 
afford a varying facilitation for the passage away of drainage water. 
Tims far for the mere percentage mechanical composition of the soil, 
based upon the arbitrary division into stones retained by a one-fourth- 
[neh sieve and fine soil. 
We have now to go into the very important question of the weights 
per acre represented by the samples taken in various years from the 
various plats and at various depths. The whole of the figures relat- 
ing to the great number of samples taken need not be given, nor 
would their record serve any useful purpose, but it appears important 
to give a resume of the hitherto unpublished data on which are based 
the acreage weights now adopted at Rothamsted, and Sir Henry Gil- 
bert lias very kindly furnished me with such a resume. 
The methods of sampling and weighing and of sifting and dividing 
the samples into stones and fine soil have already been described. 
In comparing the results obtained in the 18G5 and 1881 samples ii was 
noticed many years ago that the 1805 samples representing the first, 
inches were approximately one-tenth lighter in weight than those of 
