I first became aware of the importance of seasonal variation in the agriculture of developing countries when working in Ethiopia more than fifteen years ago. During the course of field work in the central highlands, fellow workers and I found the symptoms normally associated with famine: fires unlit when the main meal of the day should have been cooking; empty food stores; naked, pot-bellied children shivering in the chill of the monsoon. In Western countries the name of Ethiopia has since become almost synonymous with famine, yet what we then saw was not famine. Local people assured us that such things happened at that time every year. After the harvest things would greatly improve. However, it was not until I made it the subject of a special study in another part of the central highlands (Gill 1977) that I became aware of the great complexity and far reaching importance of seasonality. Many years later I was to come across similar conditions of seasonal hardship in the very different setting of Bangladesh. There in the lean season it is not at all uncommon for farm labourers and their families to eat only once in two, even three, days.

That circumstances like these are by no means exceptional is evinced by a growing volume of literature from all over the developing world on the relationship between seasonality and rural poverty there. Schofield's 1974 review of studies on seasonality of nutrition was pioneering, bringing as it did many sometimes obscure studies within a coherent framework and before a wide audience.