Key points are not available for this paper at this time.
Summary 1. The conditions which must be fulfilled in the design of an animal husbandry experiment if the results are to be statistically valid are summarised, together with possible methods of improving the efficiency by suitable design. 2. The results of a pig-feeding experiment conducted at Rothamsted are discussed. The experiment showed that green food was essential to young pigs under the conditions of the experiment, that pigs fed on a wet mash grew at a faster rate than pigs fed on dry meal, the difference being due to the greater quantity of food consumed by the former, and that the effects of numbers in a pen (giving equal floor space per pig) were negligible in spite of the greater possibilities of exercise in the larger pens.
F Yates (Mon,) studied this question.