62 SOCIAL SCIENTIST decline in land revenue burden in terms of real produce.4 A similar argument was put forward by Neil Charlesworth, who believed that expansion in cultivated area, commercialisation of crops, increase in irrigated area, better markets and rising prices accounted for increased agricultural prosperity during the late nineteenth century.5 They viewed the recurrent famines as mere climatic cataclysm and not in social terms as consequences of large-scale pauperisation resulting from depeasantisation and proletarianisation under colonial rule. The entire problem needs fresh exploration in all its complexity and diversity by taking a long-term view of the dynamics of the agrarian economy by re-examining the forces and processes at work which resulted in chronic underdevelopment in agriculture. The overall economic performance of the colonial agriculture can be judged by a close scrutiny of the crucial indicators to determine how far growth was a possibility in the colonial context and what were the limits of the possible beyond which the colonial agriculture could not and did not expand. The entire period of colonial rule was marked by a persistent demographic instability. The size of population in any region is an elemental force in determining the level of its economic activity. The increase or decrease in the number of people can be used as an indicator of performance of the economy. The population rose and fell in ebb-and-tide type of rhythms, showing little irregular increase in primitive societies, whereas in modern societies population registers a sharp and continuous rise.6 The long-term stagnation of population in pre-industrial and colonial societies was an indicator of arrested growth, and recurrent violent fluctuations were symptomatic of cyclical crisis. Till about 1921, population showed violent fluctuation in all the Deccan districts except Khandesh, without making any noticeable advance. The recurrent famines and epidemics prevented any marked increase in population. The degree of fluctuation depended upon the intensity of famines and epidemics (Table 1). It has been estimated by Sumit Guha that the population in Deccan grew at the rate of 2 to 3 per cent per annum during the first half of the nineteenth century.7 Firstly, the figures and estimates of the earlier census are incomplete and dubious, and not at all reliable for comparison. The population of villages where census was not conducted was taken to be equivalent of average density of population in other regions. In southern districts population was estimated by dividing the total revenue by per capita revenue burden. This was in no way a sound basis for calculating the population. Secondly, there is a great deal of qualitative evidence contrary to the picture presented on the basis of such estimates. For instance, according to various settlement reports and official records, during 1835-40, the Central Division districts were half-depopulated by the effects of the mismanagement and unscrupulous rapacity of revenue officials. As a result there was a general tendency towards decline in area under cultivation, fall in revenue collections, and nonpayment of land revenue dues, resulting in large outstanding balances.