Veld dominated by Hyparrhenia and other species, in a sub-tropical 38-inch rainfall area in Swaziland, was changed into a more productive sward of Eragrostis curvula and other seral grasses by the application of nitrogen, phosphorus and lime. Nitrogen had the greatest effect in terms of both herbage production and ecological change, but responses were increased by the addition of phosphorus and lime. There were indications of yield responses to zinc in the presence of lime, and to molybdenum, but trace element effects were generally small and inconsistent.

The relations between plants and their water supplies are complex and there is no uniquely critical soil water deficit (i.e. no unique stage in the drying cycle at which growth and/or commercial yield begins to be significantly affected by increasing dryness of the soil). Evidence is presented that different deficits are ‘critical’ according to the type and depth of the soil, and the adequacy of nutritional supplies.

There are now too many research papers on every subject for anybody to read. Authors should aim to ‘sell’ their papers to readers. Accessibility and readability are selling points. To get a paper read, pay special attention to the title and summary. Have an informative title which does not distract the reader with useless words and phrases; start it with the most significant phrase. Put a summary or abstract at the beginning, confine it to research results, and make it easily readable.

The problems arising in field trials in plant pathology and entomology are considered under five headings: (1) experimental design, (2) conduct of experiments, (3) records, (4) analysis and (5) interpretation. Solutions to the problems are not always given, but it is useful to recognize their existence. Examples are given from a wide range of crops

A field experiment was conducted to see if a successful crop of groundnuts can be raised in North Bihar (India) during summer in rotation with paddy. The experiment was designed to find a suitable variety, and the best time of sowing and digging. Three promising varieties of groundnut, one early-bunch and two late-spreading types, were included in the test. Emergence and flowering were rapid in the early-bunch variety as compared to the two late-spreading varieties. Though flowering was less in the bunch variety, the percentage of mature kernels was higher. The earlier the sowing and the later the digging, the greater was the pod yield in all cases. On the basis of these results and other practical considerations, the best combination seems to be to sow the early variety AK 12-24A about 15 January and to dig about 22 June.

Devices are described for calculating the available-water capacity of a soil from data on the characteristics of soil in each horizon.

Several methods of measuring and estimating the available-water capacity of soils with varying degrees of accuracy are described. The errors associated with the use of the different techniques on a wide range of soils have been calculated and suggestions are made regarding the choice of method for particular circumstances.