THK HKOADHALK WIIKAT SOILS. 
71 
The difference between plat s and 7 is almost exactly proportional 
to the annual dressings of ammonium sails applied to the two plats, 
and also proportional to the increase of crops obtained under their 
Respective influence. Bui in plat 8, where the still further increase 
m the annual dressing of ammonium salts, although it produces an 
increase in crop, does not produce a proportional increase, hut is rela- 
tively more wasteful, the quantity of unutilized nitrates in the soil is 
relatively considerably greater. It is also to be noticed that there is 
a much greater decline below the depth of the pipe drainage in plats 
•I and 7 than in the case of plat 8, where the application of ammonium 
salts is admittedly excessive. As the chalk occurs not very Par below 
the tenth depth, drainage would go <m freely below, and it is clear 
that there would be great loss by natural drainage during the winter 
months of much rain and in the absence of an actively feeding crop 
to consume the nitrates at tin* surface. It is of course this winter 
drainage that involves the greal loss of soil and manorial nitrogen 
Which necessarily OCCUrS in the growth Of w heat and similar cereal 
crops. 
The autumn application of ammonium salt>. though still experi- 
ment ally maintained on plat L5, has long been abandoned in practical 
farming. Its wastefulness La shown by the relative crop records of 
this plat, as compared with plat 7, similarly manured bul receiving 
Its ammonium salts' in the spring (see Table L). Indeed, it is largely 
the experience gained at Rot>hamsted that led to the abandonment in 
modern farming practice of the autumn application of ammonium 
salt s. 
The nitric nitrogen found in October is Lesson the plat thai receives 
its nit rogenous dressing in the autumn; but at the time of sampling 
the annual dressing of ammonium salts had not been applied, and 
much nitrification would occur subsequently to the application of the 
ammonium sails at the end of October if the rest of the autumn were 
mild, and consequently greater loss of nil rogon would occur during 
the winter rains than on the spring-dressed plat. 
It is observed at Rothamsted, when ammonium salts are applied, 
say, at the end of October, that if the drainpipes happen to run 
within a week or two of the sowing, the first collection of the drain-: 
age water will show traces of ammonia; but that if there be only a 
day or two of continuous drainage, even so late in the year, the 
ammonia often wholly disappears, the w hole of the nitrogen in the 
drainage being already converted into nil rates. Indeed, in the earlier 
years of the experiments, when the ammonium salts wore throughout 
lor the most part applied in autumn, the quantity of nitric nitrogen 
In the winter drainage of the different plats obviously bore a very 
direct relation to the quantities of ammonium salts applied. 
The application of ammonium salts is. in fact, virtually tantamount 
to an application of nitrate, and it is only because ammonium salts 
