[AssemMBLy, No. 33.] 11 
has been active and most commendable. This spirit of harmony 
and interest is deserving of record, as adding so much to the effici- 
ency of our work. Some changes have occurred during the year 
Mr. Wing resigned his position, as first assistant, in June, and C. §. 
Plumb, a graduate of the Massachusetts Agricultural College, was 
secured as his successor; Mr. Watson, our stenographer, left Janu- 
ary 1 and his place has been filled by F. E. Newton; Mr. Lovett 
resigned in September; O. E. Liess came in the spring as a student, 
and has been given the duties of an assistant, which he has faith- 
fully fullfilled; K. F. Ladd, a graduate of the Maine Agricultural 
College, came as a student October 1, and on December 1 was ap- 
pointed assistant chemist. Mr. G. W. Churchill was engaged asa 
farmer last winter, and has performed his duties faithfully and in- 
telligently, relieving the director of all cares as to the detail work of 
his department. 
The duties of an agricultural experiment station, as your director 
interprets them, combines science with practice. The most careful 
scientific work is required for the establishing of principles and the 
interpretation of laws, and accurate practical work for verifying the 
conclusions gained and for suggesting applications. The great 
want of the agriculture of to-day is the establishment of principles 
which shall serve as a safe guide for reasoning, and it is only as 
one’s views offered as prophecy is submitted to the test of close ob- 
servation by the aid of the scales, the foot-rule and the measure, 
that the deficiency of our present knowledge in the very essentials 
of our work becomes manifest. It is with the intention of exhibit- 
ing the fallacy of some very universal beliefs, that I have devoted 
some space to the examination of ‘ duplicates,” for it is evident that 
when true duplicates cannot be obtained, it is unwise to expend our 
energy in attempting work over which we can have no check. To 
attempt plat-work, until we have means of determining whether 
our results are in accord with our reasoning, or until we can pur- 
posely secure duplicates, is very apt to lead to the establishment of 
erroneous belief, and to thus work an injury not only in the present, 
but as preventive to progress. Until we can obtain closer duplicates 
than we have yet secured, we can do no better than to confine our- 
selves to interpretation in terms of greater or less, rather than in 
terms of exact bushels or pounds. Indeed, until agricultural science, 
so called, can be subjected to the tests that are recognized as essen- 
tial to correctness in other sciences, we cannot hope for that progress 
which we desire. Fortunately, there is reason to believe that a 
more careful study into agricultural problems will lead us toward 
the period when we can give value to the various factors which in- 
fluence crop production, and thus be able to secure duplicate condi- 
tions and duplicate results, through the reducing of apparent varia- 
bility to uniformity, through the logical application of corrections 
which shall be justified by a more exact knowledge than we now 
possess. These results, however, cannot be expected of the plat 
