On the Rain and Drainage - Waters at Rothamsted. 7 
or longest running does not, however, always imply the largest 
discharge. A study of the records that have been regularly 
made of the size of the streams issuing from the different pipes 
would show that the largest discharge is from Plots 13, 17, 
Z&lL Next in order stand Plots 11, 18, 12, 6, 5, which 
appear to be very equal. The third group is formed of Plots 
15, 16, 14. The fourth group consists of Plots 10, 7. Tiie 
fifth of Plots 9, 8. Last of all come Plots 2, 19. The amount of 
water passing off by the drain-pipes in the case of the fifth 
group is perhaps two-thirds of that discharged by the first group 
of plots ; this will probably represent the extreme range of 
variation, if we except Plots 2 and 19. The cause of some of 
these differences apparently admits of explanation ; thus a part 
of the water from Plots 8, 9, 10, 14, is probably intercepted by 
dells, through or alongside which the drain-pipes pass. The 
character of the subsoil is also an important factor. Where the 
subsoil is exceptionally stony, as is the case in certain places 
in the field, the water conveyed by the drain may be more or 
less lost, this loss of water being greatly facilitated by the 
character of the drain-pipes used. 
The account just given of the mode of running of the drains in 
Broadbalk Field will seem strange in many particulars to those 
who are familiar with the working of deep drains on heavy 
clay land ; the drains on such land will usually run uninter- 
ruptedly throughout the winter, while the drains in Broadbalk 
Field continue running only a few hours after rain has ceased. 
The cause of difference lies in the fact that the Broadbalk drains 
are comparatively near the surface, and that any accumulation of 
subsoil water is prevented by the chalk which underlies the soil 
at a depth of about 10 to 14 feet from the surface. The drainage- 
waters from Broadbalk are thus a discharge of the water percola- 
ting through the soil, while the drainage from the deep drains 
in heavy land is mainly supplied from a reservoir of subsoil 
water. 
We have now to describe the manures applied to the various 
plots in Broadbalk Field. The manures applied, and the average 
produce obtained during the last fifteen seasons, 1866-7 to 
1880-81, through more or less of which the investigation of the 
drainage-waters has continued, and during the 30 seasons 
(1852-81) of nearly uniform manuring year after year, will 
be found in Table XXXVII. In the case of Plots 2 and 3 
the manuring has remained unaltered since 1844, the date of the 
first experimental wheat-crop. 
The dressing of " mixed mineral manure " contains in every 
case 3^ cwts. of bone-ash superphosphate, 200 lbs. of commercial 
sulphate of potassium, 100 lbs. of sulphate of sodium (nitre- 
cake), and 100 lbs. of crystallised sulphate of magnesium per 
