The effect of sulphur fertilisation on yield and herbage composition was investigated in two successive years at two grassland sites, one with ryegrass and one with ryegrass/ clover, where the available sulphur in the soil was less than 10 mg kg-l. When supplies of N, P and K were not limiting, herbage yields increased with added sulphur at both sites, particularly at the second and third cuts and in the second year. Sulphur additions also increased the concentrations of total S and so42--s in the crop, and reduced the N: S ratio. The figures for these parameters in herbage from control plots confirmed that low yields in the presence of normal nitrogen fertilisation were probably caused by sulphur deficiency.

The effects of the application of fertilisers, especially N, and trace elements on the contents of Cu and Mo in mixed grassland herbage and its main constituent species, have been studied. Results have been obtained from field experiments on three mineral soils and on a deep acid peat, the herbage having been cut at stages of growth considered suitable for conservation as silage three or four times per year for at least three successive years.Applying N without Cu reduced the Cu content in herbage in three experiments. At these three sites, applied Cu produced slightly higher yields and a small increase in the Cu content in herbage, which was greater with than without applied N. In the fourth experiment, on a soil containing adequate Cu, applying N increased the Cu content in mixed herbage, ryegrass and cocksfoot, except with 56 kg ha-' per cut, where the content in mixed herbage was reduced slightly by the elimination of clover. Where the soil was adequately supplied with Cu, either naturally or from additions of this element, ciovers contained more Cu than grasses. The Cu content was higher in herbage from the October than the June cuts, the mid-summer cuts having intermediate contents. Cu extracted by 0 . 0 5~ EDTA from the soils agreed with the effects on herbage contents, but a higher level of extractable Cu may be required for intensive grass than for optimum cereal production.Applying Mo had no effect on yield, but produced a large increase in the Mo content in herbage on a mineral soil; the increase was greater in clover than in ryegrass. On peat, applied Mo had much smaller effects on herbage contents, giving similar increases in both clover and ryegrass. Applying N reduced the Mo content in herbage, both in the absence and presence of added Mo. The effects of the latter decreased gradually in successive years. Applying P, especially at high rates, resulted in s!ightly lower Cu and Mo contents in herbage.

The effects of the application of fertilisers, especially N and trace elements, on the content of Co in mixed grassland herbage and its main constituent species, have been studied. Data have been obtained from field experiments on three mineral soils and on a deep acid peat. The herbage was cut at the silage stage of growth three or four times a year for at least three successive years. Without added Co, N applications increased Co content in herbage on peat but reduced it on mineral soils mainly by eliminating clover which had a higher Co content than grasses. Applying 0.5 kg ha-1 Co as cobalt sulphate increased herbage Co content, the increase being greater on peat than on mineral soils. These increased levels were unaffected by N applications on mineral soils but were further increased by applied N on peat. The number of years during which the Co additions were effective in producing herbage containing at least 0.1 mg Co kg-l dry matter varied on the different soils, ranging from barely three years on one mineral soil to over five years on peat. Without added Co the last cuts in October usually had the highest Co content and the first cuts in June the lowest. With added Co, the seasonal effects were more variable, but the Co content in the herbage frequently decreased from the first to the last cuts.