20 On the Rain and Drainage - Waters at Rotliamsted. 
dischargfinsr for at least an hour when the first collection was 
made. The size of the stream at 7 A.M. was about half the 
pipe. The waters were then slightly turbid, and especially from 
Plot 13 ; the subsequent collections were all clear. During the 
first three hours it will be seen that the waters became stronger, 
but that afterwards they became decidedly weaker. We are 
here perhaps at the turning-point for the season : the layer 
of soil richest in chlorides lies now not far above the level of 
the drains. 
The runnings of these plots on June 2 and 3 have been 
already given (Table XLIII.) ; the succeeding runnings on 
July 1 and August 3, 1879, and February 17, lb80, are given ir> 
Table XLIV. These are by no means the only runnings which 
occurred during the season, which was very wet, they are se- 
lected as giving: a fair idea of the condition of the waters at 
certain characteristic periods. It will be seen that from June 
onwards the chlorides in the drainage-waters tend to increase 
as the flow of water diminishes, the upper soil being now 
poorer in chlorides than the soil immediately surrounding the 
drain-pipes. 
The nitrates, being salts nearly equally diffusible with the 
chlorides, generally rise and fall with them, though frequently in 
very different proportion. Cases, however, may occur in which 
the chlorides and nitrates are not distributed throughout the soil 
in the same manner. As nitrification takes place most actively 
in the upper layers of soil, a band of nitrates may be formed 
near the surface of a soil in which the chlorides are equally 
diffused. In such a soil the nitrates may diminish in the 
drainage-water with a diminishing flow of the drains, while the 
chlorides increase. An excellent example of this is afforded by 
the runnings of Plot 15 on Nov. 15 and IG, 1880. Plot 15 had 
received its ammonium-salts on Oct. 25 ; heavy rain followed 
from the 26th to the 29th ; the chlorides were thus washed into 
the lower layers of the soil before any considerable nitrification 
had taken place. On Nov. 15, when the drains next ran, the 
surface soil had become rich in nitrates, the chlorides occupying 
a much lower level. In three successive collections the nitrogen 
and chlorine found were, in parts per million, as follows : — 
Xitrogon ns 
Niirat«3. 
Chlorine. 
November 15, 4 p.m. 
67-8 
^9-0 j 
16,8 a.m. .. 
500 
60-6 ' 
16, 2 r.M. .. 
34-6 
6:51 
