The problems resulting from genotype × environment interactions are discussed, giving an example where the application of statistical techniques to maize breeding makes it possible to assess the effects of such interactions. A method of maize breeding is proposed which enables a number of objectives to be achieved from a single breeding programme.

Taichung (Native) No. 1 rice was grown under controlled conditions with a basic temperature regime of 30°C by day and 20°C by night. Periods of 2, 4 and 12 hours at 15°C (59°F) were imposed during darkness for the first or second 40 days, or for the first 80 days after sowing. Floral initiation was delayed by cool periods of 2 hours at night or longer, and delay from this cause was always accompanied by reduced grain yields, resulting from fewer spikelets per panicle or increased tiller sterility. This effect was not overcome when the cool periods ceased after floral initiation. Cool periods longer than 4 hours had a greater effect when they were given in the first 40 days.

Three varieties of rice, selected for their differences in height and tillering ability, were grown side by side in 10-row plots to determine the effect of varietal competition on plot yields. Plants in border rows performed differently from those in the centres of plots, higher-tillering or taller varieties tending to depress the performance of adjacent plants. Comparisons among varieties based on plot yields, determined with and without excluding the border rows, gave different results. Excluding two rows on each side of a plot was sufficient to eliminate the effects of varietal competition on yield data.

Compost or ammonium sulphate improved and maintained cotton yield for 6 years but both, together with triple superphosphate, were needed to maintain it for 9 years, after which yield declined and lime was necessary. Compost supplied necessary but insufficient phosphorus, potassium, calcium and nitrogen. Manganese toxicity was severe when potassium chloride was used on very acid soil. At suitable soil pH and with adequate phosphorus and cations, response to nitrogen was large, but for it to be most effective, calcium seemed necessary to maintain the cation: anion ratio. Without lime, yields declined with heavy rain; with it they were largest in the wettest seasons.

A rapid, simple method is described for estimating the area of rubber leaves from two measurements on the middle and one of the side leaflets. The relationship between the area of a leaflet (A) and its length × breadth (LB), described by the expression A = 0.654 LB, does not vary between the three leaflets or between leaves of different ages, and clonal differences are slight.

The decline in fertility of red ferrallitic soils in Uganda under arable cropping, and its restoration under permanent vegetation, is shown to be associated with losses and gains of major nutrients by the top-soil. A fertilizer mixture formulated to correct the losses and imitate the gains of nutrients consistently increased crop yields and maintained soil fertility over 4 years. Reasons for the success of the fertilizer mixture and its potential use in other tropical areas are discussed.

Controlling weeds until 6 weeks after crop emergence gave a yield of soyabeans only 85 per cent that of the clean-weeded control treatment. Delaying weeding until 3 or 6 weeks after emergence gave yields 89 per cent and 69 per cent respectively of the control. Narrow crop rows (30.5 cm.) gave higher yields than wide rows (61 cm.). Defoliation of soyabeans reduced pod number, seed size and yields more with late and severe defoliation, and with lower plant populations. Delayed harvesting during the dry season led to seed shedding and yield was at the maximum for only about one week.

Complete nutrient ± boron treatments were imposed on seedling and young oil palms growing in sand and water culture. A number of symptoms were observed on the boron-deficient palms, including the development of high endogenous auxin levels. The relation of morphological symptoms with IAA concentration is discussed. Comparisons were made of dry matter production, leaf area and leaf boron concentration.

Potassium application appeared to increase dry matter production and yield of oil palms, primarily by increasing leaf area, while nitrogen increased both leaf area and net assimilation rate. Responses to phosphorus and magnesium were harder to interpret. Production of vegetative dry matter attained a fairly constant level at the higher rates of fertilizer application, while the ratio of bunch yield to total dry matter production was little affected by fertilizer treatments. The importance of these results for oil palm breeding, and their possible value in diagnosing fertilizer requirements, are briefly discussed.

A previously undescribed mosaic disease was observed on groundnut plants in several districts of West Malaysia. The virus was transmissible through sap, grafting and the aphids Aphis craccivora, A. gossypi, Myzus persicae, Rhopalosiphum maidis and Toxoptera citricidus. A. craccivora, the principal vector, shows dual transmission. The host range of the virus included mostly leguminous plants such as Glycine max, Cassia occidentalis, Crotalaria anagyroides, Indigofera hirsuta, Stylosanthes gracilis and Phaseolus vulgaris, and a non-leguminous host Chenopodium amaranticolor. The longevity in vitro of the virus was between 40–48 hours, its thermal end-point between 63° and 65°C and the dilution end-point 10−4.