ECONOMIC HISTORY OR AN APOLOGIA FOR COLONIALISM 55 remarks would suffice Except in the Bengal, where the land revenue was perpetually fixed, in other areas the land revenue could be changed from time to time. The state was the effective landlord. A classic description of the ryotwari system was given by Marx: "In Madras and Bombay we have a French peasant proprietor who is at the same time a serf, and a metayer of the State. The drawbacks of all these various systems accumulate upon him without his enjoying any of their redeeming features. The Ryot is subject, like the French peasant, to the extortion of the private usurer; but he has no hereditary, no permanent title in his land, like the French peasant. Like the serf he is forced to cultivation, but he is not secured against want like the serf. Like the metayer he has to divide his produce with the State, but the State is not obliged, with regard to him, to advance the funds and the stock, as it is obliged to do with regard to the metayer.9'3 In Bengal it became evident over time that the share of the state in the appropriated surplus from land was relatively less than that of the landlord. Precisely, as Bengal remained the chief citadel of the British Indian Empire—and this for political reasons—this system was not changed. Collaboration with the landlord was an essential element of the colonial strategy, and this was also implemented in other regions. The Punjab Land Alienation Act of 1901, for instance, was implemented to protect the traditional vested interests in land against fresh incumbents.4 Again, a significant departure of the British land revenue settlements front the previous Indian regimes was assessment on the basis of what and how much the land ought to produce and not on the crop that was actually raised. The actual revenue, therefore, was often much above that under the previous regimes. In 1899-1900, e.g., land revenue contributed more4han a third of the total revenues of the government.5 Over the years, however, the income of the government was increasingly more from various cesses than from the imposed net land revenue. la Madras (excluding Malabar and South Canara), for example, a comparison between the averages for the periods 1861-64 and 1894-98 suggests that while the area under assessment had increased approximately by 28 per cent, the gross income from cesses had increased by over 700 per cent and the gross revenue demand, inclusive-of cesses, by 51 per cent.8 A high gross land revenue demand created the conditions for the penetration of moneylending capital in the agrarian sector. E. Stokes, B. B. Chaudhuri and D. Kumar fail to perceive the responsibility of the State in augmenting the domain of the usurer which made th6 feudal relations of production more complex and the agrarian economy more miserable. The really important and characteristic domain of the usurer, wrote Marx, "is the function of money as a means of payment. Every payment of money, ground rent tribute, etc., which becomes due on a particular date, carries with it the need to secure money for such a purpose."7 Besides this, the demand of the British bourgeoisie for crops