Thirty-eight experiments on swedes and turnips, using factorial designs without replication, were carried out to measure the responses and two-factor interactions produced by applying nitrogen, phosphorus and potassium. The effect of F.Y.M. was measured at fourteen of the centres.

The yield of tops was substantially increased by nitrogen. Phosphorus produced a moderate response but potassium had practically no effect on top growth.

The yield of roots showed small to moderately large responses to nitrogen in most experiments. There was a considerable variation in the rate required to produce the largest response, depending on the previous cropping of the field.

Phosphorus increased the yield of roots in all but four of the experiments and this nutrient generally had a greater effect than either nitrogen or potassium.

The response to potassium was significant in only half of the experiments and its effect on yield was normally less than that of nitrogen.

The two-factor interactions were practically without exception small and far from significant, the highest values usually occurring on responsive soils. The means show no interaction between nitrogen and phosphorus, and small positive interactions between nitrogen and potassium and between phosphorus and potassium.

F.Y.M. produced higher yields at the majority of the centres but had very little effect at some. At suboptimal P2O5 levels F.Y.M. tends to increase the response to nitrogen, but with adequate phosphorus this manure reduced the most profitable nitrogen rate. The response to phosphorus was markedly reduced by placing F.Y.M. in the bottom of the ridges but the effect of ploughing-in this manure during autumn or early winter on the phosphorus response was very small. In all experiments F.Y.M. produced a substantial reduction in the response to potassium.

As a result of the inadequate phosphorus rates these experiments do not provide satisfactory estimates of the amounts of nutrients supplied by F.Y.M. but suggest that 15 tons per acre will supply at least 15 lb. N, 15 lb. P2O5 and 70 lb K2O to the first crop grown after application.

The mineral composition of the roots showed large variations between crops grown on different soils. The application of a nutrient generally produced a small increase in its percentage in the roots. The nutrient uptakes showed a much greater apparent recovery of both nitrogen and potassium than of phosphorus.

The most profitable nutrient dressings have been calculated from these experiments, and at current prices about 65 lb. N, 125 lb. P2O5 and 125 lb. K2O should be about the average optimal rates per acre in the absence of F.Y.M. In the presence of a normal dressing of the latter suitable optimal rates seem to be about 50 lb. N, 110 lb. P2O5 and 55 lb. K2O.