34 
Beport of the Director of the 
used alike on all these plats was 400 lbs. per acre in 1882, 1883, 1884 
and 1886; 1,000 lbs. per acre in 1885, and no fertilizer in 1887. 
Conclusions. 
These trials indicate the utter unreliability of field experimentation, 
and should suffice to convince the public of the lack of certainty which 
attends all general conclusions gained by this process. I trust the 
time may arrive when this plat work instead of being forced upon 
experiment stations, will be so condemned that if used at all, it shall 
be only with the apology of a special purpose to be served, and cultivated 
fields shall be relegated to their only true purpose in experimentation, 
that of verifying results otherwise gained, or for teaching practical 
methods of application, or for purposes of illustration. For six years 
now this station has devoted much expenditure and labor to its plat 
areas, forced by a mistaken public sentiment, and a review of its work 
indicates that little or nothing has been gained in the direction of 
manurial questions, or questions of a general nature relating to the 
growing of crops. The apparatus for station research are the laboratory, 
the study, the garden, the greenhouse and fields, which offer space for 
incidental uses, and whereon experimentation can be located. The 
fallacy of equal areas permanently staked off, and with a historical 
record of treatment, is too plausible to be at once overcome, and the 
idea appeals so strongly to the half-educated or uneducated mind as 
to secure intense belief and partisan advocacy, despite the continual 
proofs of failure which are constantly found in the examination of 
published records of this sort of pseudo-scientific, so-called experiment. 
That I am not alone in these conclusions can be shown by quoting 
from Professor Dr. Paul Wagner, Director of the Agricultural- 
Experimental Station at Darmstadt, Germany. In speaking of field 
experimentation he says : " They cannot be used as a means of scientific 
investigation, because their results are not obvious. Many people have 
been led into error by assuming that general conclusions could be 
deduced from such experiments." "Manurial questions of general 
importance cannot be solved in a reliable way by field experiment to 
satisfy the requirements of agriculturists." " In every scientific experi- 
ment it is essential that all circumstances which can influence the result 
of the research must be controlled. Manurial experiments on large 
patches of ground — even if only a few hundred square yards — do not 
allow such control. The investigator cannot regulate the influences 
which affect the plants; he cannot change bad, or produce favorable 
conditions; nor can he compensate for either. He must leave his exper- 
iment to all possible chances; to the changes of weather and to various 
