JOURNAL OF AGRICULTURAL RESEARCH 3 
CLASSES OF PAPERS 
The Journal of Agricultural Research publishes three classes of 
papers: 
, In the Journal proper: 
1. Major papers: Concise but complete reports of original research. 
2. Preliminary papers reporting original research. 
In separate form: 
3. Reprints of Journal papers. 
1. A major paper should chronicle a discovery or definite advance in a 
science that directly or indirectly may be of importance to the art of 
agriculture. 
Papers which are purely of systematic and taxonomic interest should 
be published in journals which are more appropriate for such descriptions, 
reserving the Journal for accounts of research in the field of agriculture, 
as distinguished from general science. Accounts of conventional tests 
and experiments which do not involve originality in plan or method nor 
provide for the control or accurate measurement of the experimental 
conditions and which fail definitely to throw light upon the relationship 
to principles or limiting factors are not considered suitable for publication 
in the Journal. The same is true of descriptions of local variety tests 
of plants or animals, fertilizer and feeding experiments, routine analyses 
which merely contribute to the mass of similar data, elaborate historical 
reviews, and, in fact, all articles which are encyclopedic and purely 
statistical rather than reports of original studies made for the purpose 
of research. 
Long historical reviews should be reduced merely to a sufficient review 
of the literature to define clearly the problem undertaken by the author 
and to show the bearing of his contribution. Logical and concise pres¬ 
entation, freed from extraneous and confusing details, is more to be 
desired than minute and tedious notation and explanation, or studied 
originality of expression. Such an article usually will not exceed 10 or 
12 printed pages in length; and seldom, if ever, should it exceed 50 pages. 
It is seldom practicable or desirable to include all the data obtained 
in an investigation. Many of the data pertaining to intermediate steps 
and individual records of negative results are unessential to the majority 
of readers and do not justify the expense of publication. The desire of 
writers to make published articles a complete record of their studies is 
often founded on the criticism of some readers who assume that, if 
everything is not given, something of importance may have been with¬ 
held or conclusions may have been incorrectly drawn. The contention 
that the reader should be able to check up completely the author’s 
findings from the published account is generally not justified, though 
the extent to which it is desirable to report details of research is admittedly 
a difficult question, not to be determined by set rules. The reader has 
a right to essential scientific evidence, but evidently there must be a 
