Abstract

These lectures which the author had the honour of delivering under the auspices of the University are the outcome of an attempt to place before the Agricultural experimentalist the mathematical limitations that govern his 'significance,' and to explain to him in broad outline the skeleton of the statistical theory that forms the basis of judging inter-varietal differences or the differences between different treatments. To a pure mathematician, the distributions of the errors of statistics and their analytical properties form the crux of the problem, while to an experimentalist the applications of the theory with all its implications bring in not only inferences but suggest new problems and new spheres of work connected with his experimentations. For in fact in no other science are theory and practice so combined, so allied and so interdependent that most of the problems connected with agricultural experimentations have not been correctly diagnosed for want of sufficient progress in statistical ideas and statistical theory. Though indeed a high technical knowledge is necessary to study the peculiarities of the soil, to suggest ways and means for the economic use of the fertilisers, to test the different varieties in respect of the yield and to evolve strains out of them to produce the greatest effect, though these and other investigations

Keywords: