ABSTRACT
When Ronald Fisher accepted the post of statistician at Rothamsted Experimental Station in 1919, the tasks that faced him were to make what he could of a large quantity of existing data from ongoing long-term agricultural studies (one had begun in 1843!) and to try to improve the effectiveness of future field trials. Fisher later described the first of these tasks as “raking over the muck heap”; the second he approached with great vigor and enthusiasm, laying as he did so the foundations of modern experimental design and statistical analysis.
