Pasture experiments may be grazed using either variable (put-and-take) or fixed stocking rates. Considerations such as the risks of introducing bias, the reliability of the data, and the resources required for these methods of grazing are discussed, and criteria are described for choosing between the methods.

The variety Nigerian Synthetic-5 was sown on eight occasions, five in the first season and three in the second, at approximately 15-day intervals from 15 April to 15 June, and on 20 August, 4 September and 23 September. The method of seed bed preparation and time of planting both affected yields, effects that could be related to soil temperature and soil moisture reserve during crop development. Maize planted in August produced the highest yield; heaps and ridges gave the lowest yield at all planting dates, while planting on the flat or in furrows produced the highest yield.

Effects of fertilizers on harvest time and yield of cowpeas were studied on upland soils in Sierra Leone. In a 25 confounded factorial trial, on soil which had been cropped successively with 2 years of Smooth Cayenne pineapples and 14 months of cassava, the treatments did not affect harvest time significantly, but N, P and NP Mg produced very favourable effects on grain yields. In a randomized complete block experiment on soil previously cropped with two successive cassava crops, there was a significant difference in total yield of dry pods between the control and most treatments.

The mass selection reservoir (MSR) is one way to take a population approach to the breeding of sugarcane. Yield data from four successive plantings of a sugarcane MSR showed consistent superiority to a commercial clone. The problem of defining satisfactory selection indices for the MSR is discussed.

Growth of the sunflower cultivar Peredovic was examined when exposed to periodic severe moisture stress, with and without soil drench application of CCG at the two-leaf stage. The experiment was conducted in sand culture under partly controlled environment conditions. Morphological and physiological measurements were made throughout the growing period. Amongst the responses to CCC were increased stem and petiole width, and decreased length of internodes, head diameter and photosynthetic area. Moisture stress reduced photosynthetic area and the root/shoot ratio. Both CCC and moisture stress increased the number of stomata per unit area.

Tall Nigerian Farafara and a dwarf line selected from the cross Combine Kafir 60 × Farafara were compared at Samaru in eight weekly sowings in 1969, which was a wet year, and 1970 which was rather dry. Whereas the dwarf line always gave more grain yield, its advantage over Farafara, and the effect of early sowing on its yield, were greater in the wetter year. In both varieties the date of flowering seemed to be determined mainly by photoperiod, but there was also a small but clear effect of sowing date.

An experiment at Hudeiba Research Station in Northern Province for three consecutive seasons, testing the effects of form, rate and time of application of nitrogen, showed that grain yield of wheat was significantly influenced by rate and time of application but two forms of fertilizers behaved similarly. None of the factors examined affected the 1000 grain weights. A second order polynomial function, fitted to yields at different nitrogen levels to determine economic levels was quadratic, best described by Y = a + 11·809x − 0·1037 x2. When almost all relevant factors were considered, 30–40 kg. of N per acre appeared to be an optimum economic level.

Glasshouse and field trials have been used to assess the effect of applying calcitic lime and gypsum on haulm weight, pod-filling and kernel yield of groundnuts grown in soils of Northern Malawi. Incorporation of lime prior to planting and topdressing gypsum at flowering both rectified poor pod-filling and improved kernel yield. Gypsum decreased haulm weight and usually produced lower overall kernel yield than lime. The relative merits of lime and gypsum application are discussed, together with the need for supplementation by phosphorus and sulphur.

A number of crosses were made to study the mode of inheritance of flower colour and number of locules per capsule of sesame. Red flower colour was dominant to white, and 4-loculed dominant over 8-loculed pods, each of the traits being controlled by a single gene pair segregating in a 3:1 F2 ratio.

Researchers commonly use unplanted alleys to separate plots in experiments with paddy rice. In studies in the Philippines and Thailand we compared the performance of rice plants grown in previously cropped areas of a field with that of plants grown in areas which were unplanted alleys in the previous season. The latter plants had yields and values for several agronomic characters that were significantly different from plants grown in the previously planted areas, with yields ranging from 8 to 25 per cent higher.

The aim of this paper is to help biologists gain a clearer understanding of what is involved in covariance analysis. It is an exposition on the derivation of the regression coefficient in the analysis of covariance, and hence on the interpretation of the covariance adjustment. The covariance adjustment is shown to be equivalent to regressing the residuals from an analysis of variance of the original variate on the residuals from an analysis of variance of the covariate.