Evolutionary studies in the cottons began with the demonstration by Harland that true cottons could be divided into four genetically distinct species, each carrying an internally balanced genotype that was disrupted in crosses with any other. Harland's genetic classification has stood the test of time, and contributes substantially to our understanding of speciation in other genera, as for example, Phaseolus. In addition, it provides a genetic basis for the observations described by Anderson as ‘introgression’.

Studies of cottons found growing wild, and of their relationship to cultivars in the same areas, led to the hypothesis that all true cottons are cultivars, and those now found wild are escapes. This hypothesis, though it led to purposeful exploration of the distribution and relationships of the cottons, was found to be inadequate. A new approach to the relationships of the cottons became possible following Thoday's demonstration of ‘disruptive selection’ in Drosophila. The disruptive situation has been shown to obtain as between cultivars and their wild relatives in Sorghum and in a number of other crop plants.

Consideration of the time scale of agricultural evolution makes it clear that the recent origin theory is untenable, and a review of the status of the ‘wild’ cottons shows that they are adequately accounted for as weedy relatives of the cultivars, descended from common wild ancestors under disruptive selection.

It is concluded that crop plant evolution has gone on very largely within the balanced gene pools which Harland regarded as genetic species.

The responses of sprayed and unsprayed cotton to ammonium sulphate and single superphosphate at o, 125, and 250 kg./ha. were tested between 1959 and 1964, in three zones of Nigeria. Seed cotton yields of sprayed trials with both fertilizers at 0 and 250 kg./ha. averaged 534 and 768 kg. ha.−1 respectively with significant and positive linear response components in all zones, but significant negative quadratic responses to phosphate in the north and to nitrogen in the south, with a significant interaction in the south. Seed cotton yields of unsprayed trials with both fertilizers at 0 and 250 kg./ha. averaged 305 and 384 kg./ha. respectively with significant positive linear responses to phosphate in the north, to nitrogen in the south and to both in the east and no interactions.

The effects of moisture stress during various stages of the development of two varieties of upland cotton (Acala 4–42 and Deltapine Smoothleaf) were studied in two irrigation experiments in the Bet-She'an valley, where summer temperatures were high and evapotranspiration in the fully irrigated treatments was up to 8–10 mm./day in July and August. Optimum time for first irrigation in the Bet-She'an valley was two or three weeks before flowering and six or seven irrigations, if timed to prevent moisture stress, were sufficient to give high yields and good lint quality. Lint yield ranged from 270 kg./ha. in non-irrigated to 2000 kg./ha. in the fully irrigated treatments in 1966.

Climatic conditions affecting the development of CBD are assessed by measuring wetness within the tree canopy and air temperature. Saturation of the tree canopy, necessary for spore dispersal, occurs most frequently at the tops of trees and the duration of wetness permitting spore germination is most prolonged at night. Night air temperatures are closest to berry temperatures and are important in assessing infection periods. Disease development in 1968 and 1969 was related to the number of infection periods during the growing season. Polythene tree covers which kept trees sufficiently dry to stop disease development were used in determining infection at different times of the year.

In 1968–9 and 1970–1 small-scale experiments were set up on over a hundred farms in Barbados, simple enough for farmers to carry out with a minimum of supervision. Results were obtained from more than 70 in each year. NK fertilizing gave no increase in yield in 1968–9, but a significant increase of 800 lb per acre in 1969–70. In each year NPK gave a highly significant increase of 1000 lb per acre. There was also a general increase in yield with increasing rainfall up to about 40 inches during the growing season and with closer than normal spacing especially under conditions of low rainfall. Planting later than usual led to marked decreases in yield.

Yields in a trial of Upper Amazon clones, derived from high yielding Upper Amazon selections, were low in the early years. One Upper Amazon clone was significantly higher yielding than any other clone, but generally yields from the Upper Amazons were comparable to those of Local Trinitario selections included as controls. The Upper Amazon clones were highly variable in establishment, yield and losses from diseases. The relationship between the performance of some of the clones and their parent seedlings is discussed.

Nitrogen was lost through leaching, both as and , even after application of the ammoniacal form in a reduced soil zone. Under saturation the loss was more through , and under submergence through nitrogen. The loss was accelerated when all the nitrogen was applied at one time at the early stage of crop growth. Application at transplanting proved unproductive. Three split application at active vegetative, reproductive and ripening phases helped to give maximum responses in both alluvial clay and lateritic sandy clay loam soil.

The effect of covers of controlled volunteer vegetation, Pueraria phaseoloides, and clean weeding on the soil moisture content and on the succulence of sisal leaves was studied. The surface soil tended to dry out further when exposed by clean weeding, but moisture in the profile as a whole was not much depleted. The plant covers dried soils to at least 300 cm. in dry periods. Although sisal leaves had a slightly higher dry matter content in dry weather there was no evidence of competition for water between the plant covers and the sisal.

Soil properties influencing N mineralization, and hence the potential N-supplying power of 154 West Indian soils, have been studied. Mineralizable N was correlated with the activity of the soil biomass measured as extractable glucose, with exchangeable bases in base-deficient soils and with total N in soils well supplied with bases, but not with other estimates of the nature of soil organic matter (e.g. extractable chlorophyll or C : N ratio). It is suggested that cation exchange capacity and percentage base saturation may influence the N-supplying power of West Indian soils more than their total N content.

A method is described for studying the distribution of absorbing roots of fruit trees using the number of light root tips per unit of soil space as a criterion for characterizing the spatial distribution of the absorbing parts of the root systems. As examples of the use of this method, some results are shown from investigations with apple trees in South-Western Germany. They demonstrate the influence of soil-type and soil management on the distribution of absorbing roots. Striking temporal variations in the number of root tips were observed in the same tree.