On the Rain and Drainage- Waters at Rothamsted. 349 
plates would, during dry weather, be aerated from below; the 
subsoils would thus be more freely exposed to air than could 
happen in an ordinary field soil. On the other hand, the 
thorough tillage which accompanies an agricultural bare fallow 
must expose the surface-soil of the field to the action of the 
atmosphere more fully than can take place in the drain-gauge. 
The supply of oxidisable organic matter must also now, after 
ten years of exposure to the atmosphere and washing by rain, 
be considerably less in the soils of our drain-gauges than in 
a soil which has recently been manured and cropped. 
Towards the end of September 1878 the soil and subsoil of 
three plots in two of the experimental fields at Rothamsted, 
which had been left as bare fallow all the summer, were sampled, 
and the quantity of nitrates they contained determined. To 
the depth of 20 inches they contained respectively 34 lbs., 41 lbs., 
and 55 lbs. of nitrogen in the form of nitrates. The first two 
soils had received no nitrogenous manure for a great many 
years, and were probably in a state of greater exhaustion than 
the soils of the drain-gauges ; the third had received nitrogenous 
manure, but had grown two crops since its application. In all 
cases the crop preceding the fallow had been barley or wheat, 
and as these crops are known to remove nitrates very thoroughly 
from the soil, we may fairly conclude that the nitrates found 
were mainly the result of nitrification during the last twelve 
months. The amount of nitrate actually found would not, 
however, represent the whole produced, some must have been 
already lost by drainage. Indeed, during the six summer 
months of 1878, about 17 lbs, of nitrogen in the form of nitrates 
had been removed in the drainage-Avater of the 20-inch drain- 
gauge. 
With this evidence before us, we are disposed to conclude that 
nitrates equal to 50 lbs. or more of nitrogen per acre may be 
produced in a single year's bare fallow of the arable soil at 
Rothamsted. 
With regard to the loss which a soil under bare fallow would 
suffer by drainage, we must recollect, in the first place, that the 
rainfall during the four years' experiments with the drain-gauges 
just quoted was far above the average, amounting in fact to 
nearly 33 inches per annum ; the drainage has consequently 
averaged 17"281 inches per annum, while with the normal 
Rothamsted rainfall the drainage would only be 10*92 inches, 
and with a rainfall of 25 inches but 7 62 inches. With a smaller 
drainage the loss of nitrates from the soil would considerably 
diminish, though not in the same proportion as the diminution 
in the drainage, as with a less rainfall the drainage-water would 
become stronger. We have one whole year of moderate 
