62 Oh the Rain and Drainage - Waters at Rothamsted. 
observed that in the case of the two years of unusually frequent 
drainage, and when every flow from the pipes was analysed, the 
amounts of nitrogen estimated to be lost by drainage from the 
autumn-sown Plot 15 was, together with the amount in the 
crop, nearly sufficient to account for the whole estimated to be 
available within that period ; and so far as can be judged, the 
accumulation of residue within the soil would fully make up the 
deficiency. It must be borne in mind, however, that that period 
was very circumscribed. 
It is obvious that in the case of the Experimental Wheat- 
field, with not many feet of clay sub-soil, and chalk below, 
favouring natural drainage, there will be much drainage from 
the land when none flows from the pipes. In fact, none does 
flow from the pipes except under the influence of continuous or 
heavy rain. The character of the drainage collected from them 
will, therefore, depend very directly on the contents of the soil 
above their level. Now, reference to the details of the analyses 
shows that the drainage-waters contained rather less chlorine 
at the end than at the beginning of the two years for which 
the loss of nitrogen by nitrates in the collected drainage- 
waters has been estimated. It is concluded, therefore, that 
none of the chlorine which had been supplied by manure 
and rain within the period, still remained above the level 
of the drain-pipes. But when the quantity of chlorine in 
the collected drainage-waters is calculated in the same way as 
the loss of nitrates has been estimated, it is found that in most 
cases a considerable amount of the chlorine supplied, and not 
appropriated by the crop, remains unaccounted for. Further, 
comparing the results for Plots 6, 7, and 8, with increasing 
amounts of chlorine supplied in the manure, the quantity 
unaccounted for is the greater, the greater the amount supplied. 
Thus, in the case of the chlorine, which is supplied in a con- 
dition at once highly soluble and diffusible, and the distribution 
of which will be less influenced by vegetation than will that of 
the nitrates, there is a considerable amount unaccounted for in 
the crop and the collected drainage, and there is evidence indi- 
cating that it does not remain above the level of the drain-pipes. 
The conclusion is that it has passed downwards by diffusion, 
and by drainage other than through the pipes. Calculation leads 
to the conclusion that nitrates have disappeared in the same way. 
In the case of autumn sowing, the manure is on the ground 
four or five months before the commencement of vegetation, and 
during a period when there will usually be much more frequent 
drainage from the pipes than subsequently, It is to be sup- 
posed that under such circumstances a larger proportion of the 
nitrates which are formed near the surface will get directly 
