Onion bulbs were stored at 30°C at relative humidities of 50, 80 and 95 per cent in 1971–72 and at the same temperature at 35, 50, 60 and 70 per cent RH in 1972–73. Losses from rotting, sprouting and desiccation were recorded. The latter were lowest at 95 per cent RH and not correlated with humidity in the range 35–80 per cent RH. Rotting losses similarly were not correlated with humidity but were highest at the highest humidity.

Upland rice and Stylosanthes guyanensis were sown simultaneously in pure and mixed culture at varying densities at Na Pheng, central Laos. Vegetative growth of the rice was negatively related to stylo density over the range o to 81 stylo plants/m.2 Rice grain yield varied from 98·4 gm./m.2 in monoculture to 63·4 gm./m.2 when grown with 81 stylo plants/m.2 Rice was more competitive than stylo, but stylo growth was relatively independent of rice density over the range 20 to 120 rice plants/m.2

Simultaneous undersowing of upland rice with Stylosanthes guyanensis at Khon Kaen, north-east Thailand did not affect vegetative growth of rice but reduced grain yield by 19 per cent. Grain yield was not significantly affected if stylo undersowing was delayed by 10 days. Stylo growth was reduced by one half by delayed sowing and was negatively related to dry matter yield of rice variety.

The growth and nitrogen yield of Stylosanthes guyanensis undersown with upland rice were negatively related to level of urea application at Khon Kaen, north-east Thailand. In the absence of applied urea the nitrogen yield of rice monoculture was 3·3 gm./m.2 whilst the addition of stylo increased nitrogen yield of the association to 9·6 gm./m.2 Stylo was initially dominated by rice but exploited environmental growth factors more effectively in the latter part of the growing season, accentuating competition for water and phosphorus. Simultaneous stylo and rice sowing reduced rice grain yield by 12 per cent, the greatest reduction occurring at a low level of urea application, mainly due to abortion of spikelets after panicle exsertion.

Introductions were made to Oxford, England, of a number of strains of the pearl lupin (Lupinus mutabilis Sweet) from South America. In the first season of growth, two of these strains showed considerable promise for development as a field crop in England. Analysis of the seeds confirmed that they are comparable to the soya bean in their content of protein and oil. The chief disadvantage of the species is that the seeds contain toxic alkaloids. Alkaloid-free strains of this species have twice been bred in the past; the seed of these strains has been lost, but such work could be done again. (It is regretted that there is no surplus seed available for distribution from Oxford at present.)

Hybrids made for the first time on six experimental photosensitive sorghum seed parents (whose breeding is described) were tested with their pollen parents, which included good dwarf varieties, in replicated small plot trials at Samaru in 1972. Grain yields of up to 5000 kg./ha. were recorded from hybrids with acceptable plant types and grain qualities, which was 40 per cent more than from the best varieties. Some hybrids, though giving over 6000 kg./ha., were unacceptable due to excessive plant height or poor grain appearance. Seed parent yields, determined in an adjacent trial, ranged from 1700 to 2900 kg./ha.

The response of tomatoes to applied N, P and K was determined in six trials in the forest and savannah zones of Nigeria. In the savannah area, N was the most deficient nutrient, with an optimum level of 60 kg. N/ha., while P and K were required at lower rates of between 20–60 kg. P2O5 and K2O/ha. depending on initial soil nutrient status and moisture availability. In forest soils developed from sedimentary rocks, P requirement was high (120 kg. P2O5/ha.) and response to N was small, but K could be important on the well-drained soils in this area because of low K reserves (0·1 per cent total K) compared with soils from igneous rocks in the savannah with more than 0·6 per cent total K. The potential of the savannah for large-scale tomato production is discussed, as well as factors limiting yields in Western State of Nigeria.

The fifth leaf of tomato plants sampled at the early flowering stage (seventh to eighth week after sowing) gave good assessments of N, P and K requirements of the tomato crop growing in the field but total N was less useful than nitrate-nitrogen for diagnosing nitrogen needs. The levels of critical N, P and K in the crop were 0·4 per cent NO3-N, 0·4 per cent P and 4 per cent K, with severe deficiencies when levels were less than 0·1, 0·15 and 2 per cent respectively.

Genotypic correlations and path coefficients were used to determine the importance of various characters as components of yield in castorbean. Number of primary branches per plant and 100-seed weight had high positive correlations with yield per plant and with each other. Oil content was significantly and positively correlated with 100-seed weight. The path analysis revealed that number of clusters per plant, number of branches, height at maturity and yield of the first spike had the highest direct effect on yield per plant. Sixteen selection indices were constructed using these four yield components, the index combining all four being about 54 per cent more efficient than straight selection based on yield alone.

Equal volumes of irrigation water, applied in different ways to a crop of haricot beans in two seasons, showed that the best practice was 60 mm. irrigation every ten days.