With Plate 3)Experimental work showed that considerable improvements in fertilizer efficiency could be obtained for some row crops and for potatoes by adopting suitable placement techniques (Cooke & Widdowson 1953;Cooke, Jackson & Widdowson, 1954). In these earlier experiments comparisons between placing and broadcasting were made using granulated PK or NPK compound fertilizers. Therefore it was not possible to determine which of the individual plant nutrients was improved in efficiency through being placed to the side of the seed. For cereals Crowther (1945) showed that higher yields could be obtained by combine-drilling rather than by broadcasting dressings of superphosphate, but little information is available on the benefits of placing nitrogen and potassium. It appeared unlikely, however, that the gains from placement obtained in earlier work on row crops were due entirely to improvements in efficiency of phosphate in the compound fertilizers tested since most of the soils used contained satisfactory supplies of ' available ' phosphorus and responses to phosphate fertilizers were therefore unlikely. In the work described here experiments were carried out on potatoes to measure the possible gains from placing nitrogen and potassium fertilizers. In addition, comparisons between placing and broadcasting of nitrogen, phosphorus and potassium fertilizers were made on green peas, broad beans, kale and maize.
NATURE OF THE EXPERIMENTS
Experiments on potatoesSeven experiments were laid down in 1953 and one in 1954 to compare the yields given by placed and broadcast dressings of nitrogen and potassium. Ammonium sulphate at two levels (0-5 and 1-0 cwt. N/acre) was compared either broadcast over the seedbed before planting or placed in a single band 3 in. to the side of the seed. Both methods of application were tested in the presence of a placed basal dressing of 1-5 cwt. K 2 O/aere (as potassium sulphate) and 1-00 cwt. P 2 O 5 /acre (as superphosphate). (In addition, comparisons were made between single seedbed dressings of nitrogen and dressings applied partly at planting and partly as a mid-season topdressing. The results of this part of the experimental work have been published previously by Cooke, Widdowson & Wilcox (1957).) Comparisons between placed and broadcast dressings of potassium sulphate at one level (1-5 cwt. K 2 O/acre) were made in the presence of placed basal dressings of ammonium sulphate (1-00 cwt. N/acre) and superphosphate (1-00 cwt. P 2 O 5 /acre). In addition, two plots receiving no nitrogen and one plot receiving no potassium were included; the ten experimental treatments were arranged in a randomized block. There were four blocks at each centre except Thrales End and Hillesden where there were only three. Each plot was 22 yards long and contained 4 rows spaced 28 in. apart (-^j acre).In 1954 and 1955 factorial experiments were layed down at Rothamsted to compare all combinations of ammonium sulphate (supplying 0-0, 0-5 and 1-0 cwt. N/acre) and potassium sulphate (supplying 0-0, 0-75 and 1-5 ...