ABSTRACT
Until the eighteenth century when societies were still essentially agricultural and dominated by the never-ending problem of food supplies, there was an intimate link, which is now a thing of the past, between history and climate. Unfortunately these traditional societies have left us almost no sustained quantitative or homogeneous records of temperature and rainfall. Because of this lack of any first-hand written records the discussion of meteorological fluctuations and their effects on economic history has frequently been ill founded and very much confused. This article is therefore very much concerned with methodology; its aim is not to establish any complete solution, but rather to suggest avenues of approach.
