— 100 — single plant, since each one of the pair produces a little more than half the crop which a single plant would give and not one half exactly."* A study of the averages returned by each spacing {Vide Chart No. 6) leads to the conclusion that when wider spacings than 100 by 45 are adopted, the yield depends upon the number of plants developing per feddan, while the yields from narrower spacings are reduced by root interference. In fact the influence of root interference on the yield is the chief problem which this experiment is designed to solve.

It seems most probable that in the Sudan the cotton plant requires a larger area from which to draw the water and plant food necessary to its growth than in Egypt, where the best return is obtained from a spacing giving 14,000 hills to the feddan.I This is probably due to the dearth of nitrogenous matter in Sudan soils, and the greater demand for water made by the plant.

The protection afforded by the extra shade and support of the closergrown cotton does not influence the yield. During the earlier growth, when the plant is taking all its water from the first foot of the soil, the leaf area is not suSicient to extend any appreciable shade to the ground and consequently evaporation is high and the drying out of the soil is rapid. Later, when the plant is throwing an increased amount of shade, the root system has penetrated 3 or 4 feet below the surface of the soil and is drawing its water from a level practically unaffected by the sun and wind. " On the other hand, the free circulation of wind and the hot, dry surface soil around wide-sown plants increase their transpiration, but the self-regulation of the plant, as between the function of root and shoot, tends to minimise these differences. Thus wilting and stomatal closure on a hot afternoon appear to set in at nearly the same time in all spacings."J Arrangement as distinct from spacing, must also play a part in determining productiveness. Thus, cotton sown 80 by 60 did not yield as highly as that sown 90 by 30, although the area per hill afforded by the former method is greater. This result is probably due to interference with the roots by horse-hoes during the early stages. Eighty cm. between ridges is the minimum distance which will permit employment of mechanical methods of cultivation with any degree of safety to the plant.

With the one-metre ridge not only is cultivation facilitated, but irrigation is carried out more rapidly and with less waste. The cotton soil of the Sudan cracks very badly as soon as it dries, and as small ridges dry up sooner and develop bigger cracks than large ones, they are much more liable to be washed away. The result of this is a waste of water and the loss of plants which fall down when the earth supporting them is carried away.

* "Analysis of Agricultural Yields," Part I. Balls and Holton.

t Ibid, t Ibid.