In a paper read before the Royal Statistical Society in 1934 (1) appears the following quotation of 'an eminent Professor of Agriculture of an earlier date "Damn ;h;luplicate plot; give me one plot and I know where This certamly showed a realisation of the variability of plant 'material, but was also a confession of inability to overcome it and obtain .valid experimental results. The present paper attempts to show to -what extent the statistician has been able to provide adequate answers to the difficulties of dealing with the variable material in pasture trials. Much of what is presented here -applies to other branches of science, and many of the principles involved are adapted from quite different branches.
