for Ttcenty Years in succession on the same Land. 
317 
that there was, in ahnost every year of the eight, comparatively 
little clifFerence in the amount of ammonia required to yield a 
bushel of increase of barley on Plots 2 and 4 of Series II. with 
superphosphate but without the silicate, and on Plots 2 and 4 of 
Series III. with superphosphate and with silicate. On the other 
hand, on Plots 1 and 3 of Series III., without superphosphate, but 
with silicate, the amount of ammonia required for a given effect 
was much less than on the corresponding plots of Series II. without 
the silicate. There was also a greater increase of straw by the 
use of the silicates where superphosphate was not, than where it 
was employed. 
The next experiments to consider are those with nitrate of 
soda alone (Series IV.). 1 N received, for nineteen years in 
succession, 275 lbs. nitrate of soda, containing nitrogen = .50 lbs. 
ammonia ; and 2 N received, for the first five of the nineteen years, 
double the amount, or 550 lbs. = 100 lbs. ammonia, and afterwards, 
for fourteen years, only 275 lbs., as Plot 1 N. But as, in the 
first year of the twenty, both plots received superphosphate of 
lime and sulphate of potass in considerable amount, which 
doubtless increased the effects of the nitrogen subsequently sup- 
plied for many years, if not for the whole period, the results 
of I N are not strictly comparable with those of 1 A receiving 
annually the same amount of nitrogen as ammonia-salts, nor are 
those of 2 N comparable with those of 1 A A. As the figures 
stand, however, the average of twenty years with ammonia-salts, 
and of nineteen with nitrate of soda = 50 lbs. of ammonia, shows 
with the ammonia-salts 3'68 ammonia, and with the nitrate, 
nitrogen = only 2'74 lbs,, required to yield 1 bushel increase of 
grain and its straw ; and with the double amount during the 
first few years, the ammonia-salts show 3"53, and the nitrate only 
2 81 lbs. required. It has already been explained (pp. 290-2) that 
enough phosphoric acid and potass were applied on the nitrate 
plots in the first year, to supply as much of these constituents as 
would be contained in the excess of produce by the nitrate over 
that by the ammonia-salts throughout the subsequent period ; 
so that, obviously, only part of the better result of the nitrate 
can be supposed to be due to the condition of combination of its 
nitrogen. 
The result is, at any rate, remarkable, that after mineral manures 
once applied, nitrate of soda alone should, foi nineteen years in 
succession, yield a result in proportion to its nitrogen, compara- 
tively so little inferior to ammonia-salts used every year in con- 
junction with superphosphate, or with superphosphate and salts 
of potass, soda and magnesia. 
The next experiments are those of Series V., in which rape-cake 
was used without, and with mineral manures. During the first 
