1. A small-scale plot experiment which had been conducted from 1947 to 1952 to study the effect of massive dressings of fertilizer nitrogen, with and without phosphate and potash, on the yield and quality of a rye-grass dominant sward was continued from 1953 to 1958 with two changes in the mineral treatments in 1953. A 4 × 4 Graeco-Latin square was used.

The nitrogen treatments applied each year were: (1) no nitrogenous fertilizer, (2) 260 lb. nitrogen, (3) 520 lb. nitrogen and (4) 416 lb. nitrogen per acre per year.

The mineral treatments were: (A) no mineral fertilizer, (B) 336–504 lb. K2O per acre per year depending on nitrogen treatment, (C) 101–202 lb. K2O and 81–101 lb. P2O5 per acre per year depending on nitrogen treatment, and (D) no mineral fertilizer after 1953, although K2O and P2O5 had been applied during the previous 6-year period.

All the fertilizers were applied in four or five equal dressings each year, one for each cut.

Treatments 1 and 4 were cut four times and treatments 2 and 3 five times per year.

2. Applications of phosphate did not affect the yields of herbage dry matter or crude protein, but yields were severely depressed in the absence of potash. With adequate potash the yields under all nitrogen treatments were maintained for the 6-year period and were at the same level as in the previous 6-year period. The mean yields of dry matter for the 6-year period when potash was applied were 5200, 8420, 10,180 and 9700 lb. per acre per year for treatments, 1, 2, 3 and 4. Without potash the corresponding mean yields were 5460, 6410, 6210 and 6880 lb. Average crude protein yields were 760, 1360, 1940 and 1660 lb. per acre per year with potash and 780, 1100, 1220 and 13201b. without potash.

3. The weighted mean contents of crude protein for the 6-year period averaged over all the mineral treatments were 14·4, 16·7, 19·4 and 18·2% for nitrogen treatments 1, 2, 3 and 4. The mean crudeprotein contents of treatments with and without potash were 16·7 and 17·6%, respectively.

4. The average response to nitrogen was 15·0, 10·9 and 12·4 lb. dry matter per 1 lb. of nitrogen applied on treatments 2, 3 and 4, respectively, compared with a no-clover sward, and 7·8, 7·3 and 7·9 lb. compared with a grass + clover sward. The recovery of nitrogen in terms of crude protein was 45, 40 and 40% for treatments 2, 3 and 4, respectively, when compared with a no-clover sward.

5. The botanical composition of the plots continued to be determined by the nitrogen and potash treatments. On plots receiving ‘Nitro-Chalk’ clover was virtually non-existent, whereas on plots which received potash and no nitrogen, clover was present. With a combination of potash and nitrogen vigorous swards of timothy were found but if only nitrogen and no potash was applied, poorer grasses, such as Festuca rubra, were dominant.

6. Soil analyses showed increased pH values and decreased contents of Mg with increasing applications of ‘Nitro-Chalk’. P2O5 values were low except where phosphate was being applied. The most marked change was in the K2O values which varied from 4 units on treatments A and D which received no minerals to 40 units on treatment B which received the heavy dressings of potash.

7. Mineral analyses of the herbage for 1953 and 1958 are given. High K contents were found on treatments where potash was applied and this was associated with decreased Mg and Na contents.

8. Some details are given of a field-scale trial on the Institute farm where high levels of nitrogen and potash have been applied for 11 and 12 years with successful results.