1. Twenty-two experiments on main-crop potatoes in 1955–7 compared calcium nitrate and ammonium sulphate applied to the seedbeds before planting. Urea was also tested in fifteen of the experiments in 1956–7. The three fertilizers were compared at rates that supplied 0·5, 1·Oand 1·5 cwt. N/acre.

2. Seven experiments on kale in 1955–8 compared seedbed dressings of calcium nitrate and ammonium sulphate. Urea “was also tested in. five of the experiments in 1956–8. The rates used were 0·6, 1·2 and 1·8 cwt. N/acre in 1955 and 1·0 and 2·0 cwt. N/acre subsequently. Single dressings applied to the seedbed were also compared with divided dressings applied partly to the seedbed and partly in mid-season.

3. Generally the potatoes gave only small responses to N; there was no worthwhile gain from applying more than 1·0 cwt. N/acre in 1955 and 0·5 cwt. N in 1956 and 1957. Nitrogen gave much larger proportionate increases in yields of kale, at several of the centres crops continued to respond up to dressings of 2·0 cwt. N/acre.

4. Calcium nitrate generally gave lower yields of potatoes than ammonium sulphate and the superiority of ammonium sulphate was greater with the higher dressings. Losses in yield with the nitrate were usually associated with a severe check to early growth occurring in dry springs and with dressings concentrated close to the seed, but lower yields were also obtained at some centres where early growth was not damaged. Calcium nitrate and ammonium sulphate were roughly equivalent for kale when the nitrate did not damage germination; heavy seedbed dressings of the nitrate reduced kale plant numbers severely unless there was adequate rain after sowing.

5. Granulated urea tested in 1956 contained 4·5% of biuret, it delayed emergence and reduced plant establishment seriously in several of the potato experiments; damage increased with the level of manuring and was accentuated when dressings were broadcast over furrows before hand-planting. Yields given by this batch of urea were less than with other N fertilizers and when 1·5 cwt. N/acre was applied they were less than with no nitrogen at all. Purer crystalline urea (having less than 1% of biuret) used in the 1957 experiments caused no damage to emergence of potatoes and gave yields similar to those obtained with ammonium sulphate. There were similar effects in the kale experiments; granulated urea containing much biuret damaged germination severely in 1956; the purer products containing little biuret used in later years reduced plant numbers at one centre in 1957 and at the single 1958 centre. Where there was no damage to the establishment of kale, urea and ammonium sulphate gave similar yields.

6. Dressings applied partly to the seedbed and partly in mid-season gave slightly higher average yields of kale than an equivalent total amount of nitrogen all applied before sowing.