The arrangement of field experiments

Author's note. From about 1923 onwards the Statistical Department at Rothamsted had been much concerned with the precision of field experiments in agriculture, and with modifications in their design, having the dual aim of increasing the precision and of providing a valid estimate of error. These two desiderata had been somewhat confused in the minds of the experimenters, and the present paper was the author's first attempt at setting out the rational principles on which he might proceed. The paper is a precursor to the book on the Design of Experiments published nine years later.