The object of this experiment was, in the first place, to see whether, with cotton experiments, there was an advantage in planting a larger number of seeds per hill than the customary number of three; and, in the second place, to see the effect of the interaction of time of thinning with varying numbers of seeds per hill.

In cotton breeding work, where trials of new strains have to be carried out as soon as sufficient seed is available, it is important to conserve seed in all possible ways. At the same time it is necessary to plant sufficient seed to give a full and even stand for all strains.

In experiments carried out in 1936–7 and 1937–8 stand counts, made soon after germination, showed the advantage of the higher seed rates in obtaining a quick and full stand. Later counts and final counts at harvest showed a considerable evening up, although the two-seed treatment proved unreliable, giving the lowest stand in both seasons.

Plant height and weight records, made during both seasons, showed that plants from the larger hills were drawn up much more rapidly than plants from the smaller hills. This rapid elongation in stem height proved, from weight figures, to be at the expense of lateral development, weak and leggy plants being produced in these hills. The ten-seed treatment showed up particularly badly in this respect, the yield figures showing that the plants never recovered from this early deleterious effect.

The six-seed treatment, whilst giving a quick and excellent stand suffered to a certain extent from the same defects as the ten-seed treatment. If thinned early they tended to fill out and become more comparable with plants from the three-seed treatment, but when thinning was delayed they never caught up and final yield suffered adversely.