AGRICULTURAL PROGRESS AND EXPERIMENTS. 577 
the results of practice are founded may furnish us with 
valuable guidance, and is likely to prevent us from proceed- 
ing on lines which can only lead to failure; but more than 
that we cannot expect from it. The only plan which 
remains is the old one of actual experimental trial—of trial 
and error: we must make the best use we can of our 
knowledge of principles for planning our experiments, and 
when they fail, for finding out the causes of our failures, 
which we must go on eliminating until we have succeeded. 
Instead of depending, in fact, as we seem to have been doing 
-hitherto, for progress in the solution of our problem on such 
light as studies of the soil in the laboratory, made by a few 
independent specialists, may throw on matters connected 
with the subject, the time has now come, I think, for 
attacking the problem directly in the field, and making its 
solution the immediate object of systematic and continuous 
work, and of comparative experiments. Our laboratory 
investigations in connection with the soil ought, I think, 
no longer to be of an abstract character, and entirely de- 
tached from or only remotely connected with our problem, 
but be made in direct connection with it, and for the express 
purpose of throwing light on or helping us over difficulties 
as they are met with in our experimental efforts to solve it. 
By following this course, each discovery made in the labora- 
tory, instead of remaining for an indefinite length of time 
of probably little practical value, will be likely to receive at 
once its most important practical application, and be the 
means of helping on a step forward the solution of this, the 
most important problem of agricultural practice. This 
method of solving our problem is, of course, likely to prove 
a long one and difficult; but it is the most practical and 
the most certain of success, if success be not absolutely 
unattainable. It is, in fact, the well-beaten path to 
progress. In this case, the gain which success would bring 
is important enough to cause the effort to be well worth 
making, however small the chance may be. 
It may be, and it is quite likely, that-success will not be 
due to any one striking discovery, but rather to the more 
exact knowledge which the making of systematic experi- 
ments will give us of the effects of different methods of 
cultivation. Even if the measure of success we meet with 
be only moderate, while such success will be valuable for 
itself, the experiments by means of which it is gained, from 
being comparative, will have great value independently, on 
account of the greater and broader knowledge of agricul- 
tural operations, and on account of the collection of experi- 
mental facts they will leave us possessed of. We are 
N2 
