Experiments with different Top-Dressinc/i; upon IF/ieaf. 29 
not shut our eyes, for a faitlilull^-rccordcd field-ex j)crimpnt, 
though it may not fully decide the question for which it was 
instituted, nevertheless is frequently useful in other respects, and 
at all events never mischievous in its practical bearing, like ex- 
periments which have been cooked so as to suit certain purposes, 
or to support a fa\ ()urite theory. 
Amongst the anomalous results in the preceding table may be 
noticed the large increase in corn and straw obtained by sulphate 
of ammonia. This increase is very much larger than that which 
was realized bv its use in the preceding year, and likewise much 
larger than the increase obtained by Peruvian guano. Indeed 
the less favourable result which I'eruvian guano appears to have 
produced in comparison Avith its effects uj>on the experimental 
wheat-crop in the preceding year, is one of the most remarkable 
of these anomalies. 
It is difficult, if not impossible, to recognise a reason why in 
one year guano should give a much more favourable result than 
sulphate of ammonia, and in the next the latter should beat the 
former by several bushels. We cannot attribute this variation to 
difference of soil, as the experimental field in 1860 resembled 
intimately in composition and general character that on which 
the wheat experiments were performed in 1861. I can find no 
other solution for these and other difficulties and anomalies than by 
assuming that either the wheat-plant was not uniform in the experi- 
mental plots, or that the soil varied in depth and in its physicai 
character, so far as this is affected by the nature of the subsoil. 
•- I have good reasons to believe that the soil indeed varies in 
depth in different parts of the field. As the subsoil is retentive, 
the surface on the more shallow parts of the field in wet seasons 
often will remain soaked with water, when in deeper places the 
excess of water can percolate to a greater depth before it is 
arrested by the subsoil. A larger portion of cultivated soil 
thus is left in a more perfectly drained condition, than on parts of 
the field where a retentive clay subsoil comes nearer to the 
surface. Where such inequalities in the depth of the soil exist, 
and where the subsoil is of a close, retentive character, the culti- 
vated portion of the soil must be much warmer in some places 
than in others. Under such circumstances field-experiments 
cannot furnish perfectly uniform results. 
The plant, moreover, on this field was not so uniform as I could 
have wished, affording in itself a strong indication of inequalities 
in the depth or character of the soil. Indeed the produce of a 
field when ascertained on several separate accurately-measured 
plots, say of |^ or ^ acre each, is the best practical test I know for 
ascertaining whether a field is uniform in its character or not. 
Although the wheat experiments are vitiated to some cxtc nt by 
