The independent pioneer work of Baker and Stamp resulted in a remarkably close agreement on a working regional division of India; 1 and the more detailed break-down here presented will show a strong family likeness to these early efforts. Agreement on the main outlines is perhaps not surprising, since the broad fundamentals are clear enough, but in detail the problems are by no means simple, especially if a strict delimitation based on one master-principle is attempted. Indeed, this is an unattainable ideal. Structure is perhaps the most obvious guide, at any rate on the macro-scale; but it is clear that in an area like the Indo-Gangetic Plains landforms are of little help, since the divisions are as a rule either too broad or too much a matter of local detail, such as the distinction between bet and doab, khadar and bhangar; these may be of great importance in local life, but are in practice useless as bases for regional description on a subcontinental scale – and in descriptive writing, scale is of the essence. Yet the physical landscape and the life of Sind are obviously greatly different from those of Bengal, imperceptible as may be the gradations between these two extremes. In such a vast area, some sort of division other than micro-morphological must be made, and in the absence of strong physical lines climate may be accepted as a guide, though obviously in an open plain climatic transitions are also gradual.