106 Compensation for Unexhausted Manuriai Values, 
Compensation for Loss of Potash. 
We have, however, considered it desirable to introduce a 
change of another kind in our Tables as regards the compen- 
sation to be given for potash. Previously we had reckoned the 
whole of the potash to be recovered in the manure. This, 
however, we now think to be hardly correct, and that, as with 
phosphoric acid, so with the potash, there must be a certain 
loss of potash, more especially in respect of the loss by drain- 
age of the liquid portions of manure. Accordingly, we have, 
in our new Tables, suggested that three-quarters of the potash, 
and not the full amount, be credited to the manure. 
Period over which Compensation should be spread. 
In 1902 we proposed that compensation should be spread 
over a period of four years. This was based on the fact that, 
taking a rotation course, as exemplified in the Rothamsted and 
Woburn Experiments, there was evidence of the influence of 
manuriai application lasting up to the fourth crop of a rotation. 
Indeed, it could be shown by the Woburn Experiments on the 
continuous growth of wheat and barley, that the applications 
of farmyard manure made from purchased foods continued to 
show a residue from their application for a much longer period 
than this. Some influence could even be shown for so long 
a time as twenty years, the land being practically raised 
permanently in fertility, for the crops grown subsequently 
never went down to the level of the unmanured plots. 
On the other hand, we have had to take into account more 
particularly the fact that the nitrogen contained in purchased 
feeding stuffs (and it is chiefly nitrogren for which compensa- 
tion is paid) is, in the main, contained in digestible compounds, 
and is therefore excreted from the animal as urea. This uiea 
passes rapidly into ammonia, which is not only subject to loss 
in the manure, but also exerts its effect on the first and second 
crops only that are grown with the manure. The experiments 
at Rothamsted, to be described later, show practically no 
returns in the third and later crops for the inci eased nitrogen 
in the dung that is brought about by cake feeding, &c. The 
long continued effect of dung is due to the more slowly acting 
compounds of nitrogen contributed by the litter and the 
undigested residues of the food. 
In our former paper (1902) we gave instances from rotation 
experiments conducted both at Rothamsted and at ^Vobum, the 
general outcome of which was to show that when a loot-crop 
was fed on the land, and was followed by barley, the barley 
crop was materially benefited thereby ; that the next crop 
showed also a gain, but a much reduced one, and that when 
the wheat crop — the fourth in the rotation was reached, the 
