Why the thoroughness and durability of the change in the system as a whole? The organizational and military capabilities of both the Mughals and the British are certainly important factors. But these capabilities might have been of little avail if there had been no appreciable alterations in the ecological situation render- ing it significantly different from that of the Early Medieval and Sultanate Periods. Again, in the absence of precise data we shall merely speculate on what the altera- tions may have been. First, one might suggest that after a period of infilling and a growing density of settlement on the better agricultural lands a new assault was made in late Sultanate and early Mughal times on many forested areas of modest productivity and that in the process many previously detached islands of settle- ment coalesced into ribbons, or perhaps even blocks, of more or less continuous open agricultural landscape. Along these ribbons and within these blocks, the ar- teries of transportation quite likely were sufficiently improved to allow an appre- ciable augmentation in the flow of trade and communications. Although armies remained large and the local yeomanry continued to be disposed to fend off for- eign invasion, invading armies and more particularly armies of occupation no longer had to rely on the resources of a narrowly circumscribed and isolated island of sedentary settlement to sustain them. Rather, they could strike out in search of plunder and provisions in any of several directions and find what they sought at no great distance from their principal base of operation. Part of the enhanced capa- bility of armies to sustain themselves on supplies derived in the field was, of course, attributable to the physical improvement of the communications network instituted under the aegis of the Mughals and of the British themselves.

It is obvious from figures 14.4 and 14.5 that the existence of what we have termed a pan-Indian state, even for decades on end, does not preclude the concur- rent existence of a number of lesser states. Remarkably, however, before the late 16th century, when Bijapur and Golkonda coexisted for some time with the Mughal Empire, there is only a single decade characterized by the simultaneous presence on the Indian subcontinent of a pan-Indian state and one or more supra-regional states. In several instances it may be seen that a supra-regional state attains pan- Indian status only upon the elimination as a major power of the only other supra- regional state capable of opposing it. One is led to suppose that, once it has attained pan-Indian status, a state feels compelled to eliminate or suppress any would-be rivals. Such potential rivals would, in the nature of things, be states aspiring at the outset to supra-regional status. On the other hand, states of a relatively limited ex- tent and insignificant power could well be tolerated by a pan-Indian power, to whom they would have posed no real threat. Many such third-rate states have arisen, unpredictably, over much of Indian history in response to vigorous and am- bitious local leadership, only to fade rapidly into oblivion on the death of their founders. For a truly great power to have adopted a policy of systematically sup- pressing all such parvenu rajadoms or chieftainships would often have entailed far more in cost than the benefits likely to be returned. Finally, we may note that many small and medium-sized states have survived over much of Indian history merely by virtue of their remoteness from the contemporary principal centers of political and military power. This is, of course, what accounts for the continuation of Bija- pur and Golkonda as independent entities well beyond the time when the Mughals attained pan-Indian status. Their ultimate extinction, however, became a persistent goal of Mughal policy in the Deccan.

The nearly complete hegemony of the Indian subcontinent that the Mughals attained during Aurangzīb's reign (see figure 14.5) was, for reasons that were not primarily ecological in nature, never effectively consolidated. Still, one can argue that in subjugating the last significant Deccani states, the overextended Mughal power became unable to counter effectively the rising military threat of the Ma- rathas and subsequently to maintain the allegiance of ambitious provincial gover- nors, especially those on the periphery of the empire. Nevertheless, the stage for attaining unified rule over the whole of India had effectively been set. And for some decades it appeared that the Marathas were destined to assume the imperial mantle of their erstwhile Mughal overlords. As matters turned out, the task of forging unified rule over the whole of India was completed instead under the alien banner of the British. So far as the area of our analysis is concerned, this was achieved by 1848. The political unity of India was to endure for a century there- after. The dramatic alterations that the Mughals and the British wrought in the power systems of the subcontinent are quantitatively and graphically expressed in tables 14.2 and 14.3 and figures 14.7–9.

Our discussion up to this point has focused on the Indian subcontinent as a whole. Figure 14.4 and table 14.4, however, indicate that at all periods of Indian history there are striking differences from one region to another in respect to the rise of supra-regional and pan-Indian powers. To a certain extent these differences are an artifact of the method of regional analysis.¹⁸ To a much greater extent they are real. A glance at figure 14.3 should suffice to substantiate this assertion.

In broadest terms, what is perhaps most striking about the pattern of figure 14.3 is the prominence thereon of power centers on the North Indian Plain, from Gan- dhāra at one extremity to Bengal near the other. Of sixty-three powers represented on the map and named on figure 14.4 twenty-eight had their principal centers on the plain. If we consider only those nine powers that achieved pan-Indian status, we may note that seven were centered on the plain; and if we count the total num- ber of decades during which some pan-Indian power existed, we find that all but five out of sixty-two were accounted for by powers based on the plain. Control over much or all of the North Indian Plain, then, clearly affords the surest regional basis for the attainment of undisputed hegemony in the Indian subcontinent. The explanation is not far to seek. No other comparably large area of South Asia affords the ease of movement and hence the opportunities for rapid conquest and integra- tion that are encountered on the Indo-Gangetic Plain. None has a comparable pop- ulation base or, as a corollary, comparable agricultural productivity.

But the North Indian Plain is not all of a piece. Power centers thereon are not evenly dispersed but repeatedly arise in a relatively small number of localities, of which the areas of Delhi-Agra, Magadha, Kāśī, Gandhāra, and Kānyakubja, in that order of importance, are most prominent. Of considerable interest in viewing these localities is the westward shift of power away from its principal ancient base in Magadha through Kāśī and Kānyakubja, and then, after a long and puzzling temporal hiatus, to the Delhi-Agra area, which has been the prime hub of Indian political activity for most of the past four and a half centuries.

Let us now look in somewhat greater detail at the regional distribution of major Indian states from earliest historical times up to the present, beginning with the singularly important case of Magadha, the original core of expansion in the North Indian Plain. Without interruption, for nearly three and a half centuries during the first half of the Ancient Period, this area remained the seat of either a supra-regional or a pan-Indian power, and during this time no other part of the Indian subconti- nent gave rise to even a single supra-regional state. We have already touched on the

Table 14.4. Aggregate Duration of Powers with Pan-Indian or Supra-Regional Status, by Major Regions of the Indian Subcontinent and Major Periods of Indian History.

(Figures in a columns represent aggregate numbers of decades at power status indicated, allowing double counting when more than a single power is represented in a single decade.¹ Figures in b columns are ratios of figures in a columns to total number of decades in the periods specified, expressed as percentages. Because of double-counting these figures may exceed 100%.)