436 
Afjriciiltural Chemistry. 
rather proportional to the supphj of ammonia^'' would be conclu- 
sions from these simple facts, far more logical than many which 
we are expected to receive without any facts at all. 
To proceed — May it not be said that the effect of the farm- 
yard manure upon the increased growth of wheat, was due 
in some way or other to its carbonaceous substance, rather than 
chiefly to its nitrogen, and to a small extent to its minerals 
in connexion with that nitrogen ? This supposition is nega- 
tived by a comparison of the results of Series 7 with those 
already discussed, which throws some light upon the relative 
effects of the nitrogenous, the carbonaceous, and the mineral 
constituents of manure, upon the increased growth of wheat, 
on any corn-exhausted soil. Series 7 gives the average of 12 
cases, distributed over six years, in which rape-cake, either 
with or v;ithout ammonia-salts, was employed in such quan- 
tity as to provide nitrogen exactly equal to the ammonia- 
salts of Series 6 ; and as in the latter, so also in Series 7, 
minerals were always added. We have here the striking result, 
of an average difference of produce of only 9 lbs. between Series 6 
and Series 7, there being in the two cases an identity in amount 
of nitrogenous supply ; and this, notwithstanding that the rape- 
cake employed in Series 7 would supply a considerable amount 
of additional minerals, and a large quantity of carbonaceous sub- 
stance, neither of which have given an appreciable increase be- 
yond that of the equivalent nitrogen and minerals of Series 6, 
In Series 8 we have, together with the minerals, less ammonia- 
salts than in Series 6 ; and with this less amount of ammonia or 
nitrogen, we have only 1671 lbs. of increased produce, instead of 
2667 lbs. as in Series 6. In Series 9, on the other hand, we have 
witli the minerals, moi'e ammonia than in Series 6, and also a 
considerably increased amount of produce ; tliat is to say, when 
with the minerals we have applied a larger amount of ammonia in 
the manure than the " standard]," we find an average annual in- 
crease of total produce per acre of 3581 lbs. ; with only the 
" standard " amount of ammonia, an increase of 2667 lbs. ; and 
with less ammonia than " standai'd," an increase of only 1671 lbs. 
And, fmally, in Series 10, with minerals, rape-cake, and am- 
monia, containing together rnore nitrogen than in Series 6 and 7, 
Ave have, taking an average of 22 cases, and extending over a 
period of five years, an average increase of 3152 lbs. ; instead of 
2667 lbs. and 2676 lbs. in Series 6 and 7 respectively. 
Surely in the results of this comprehensive summary of ex- 
perimental evidence, we have good grounds for ccmcluding, that 
this practically corn-exhausted field was in a condition to test the 
nature of that exhaustion, and of the constituents requisite to 
restore it to that condition of practical corn productiveness, 
