Secrets of the water fort
Ya apo divya uta va sravanti khanitrima uta va yah svayamjah samudrartha yah suchayah pavakasta apo deviriha mamavantu
Apah Suktam, Rigveda, VII. 59.2
"May the waters that are in the sky, or those that flow (on the earth), those (whose channels) have been dug,* or those that have sprung up spontaneously, and that seek the ocean, all pure and purifying, may those divine waters protect me here (on earth)."
H H Wilson, Rigveda-Samhita, Vol IV, 1978, 333
IN THE small, sleepy village of Dholavira, ensconced in a corner of the isolated island of Khadir off the Gujarat coast, clinching evidence has been found to establish that the Harappans of the 3rd millennium BC had developed an amazingly advanced hydraulic system.
This is a find that could up-end established notions of how the Harappans used water for agriculture. So far, it had been commonly believed that the urbanisation of the advanced Harappan civilisation came about only on the strength of enormous food production. There must have been largescale agriculture to sustain an urban settlement which included non-agriculturists but until recently little was known about the Harappan irrigation systems.
The greater part of the Harappan empire was spread over a vast area of about 12,000 sq km and included the alluvial plains watered by the Indus and other rivers. Barring the plains along the Himalayas, the Harappan empire was largely arid or semi-arid, receiving marginal rainfall in both summer and winter.
Although perennial rivers like the Indus, Jhelum, Chenab, Ravi and Sutlej flowed through the region, the Harappans found it difficult to harness water from them because they lacked the wherewithal to make dependable cementing material on a large scale for conduits. At best, they could have tapped the floodwaters that surged into inundation channels and natural or artificial reservoirs.
Harappan methods The cultivators among the Harappans may have adopted some or all of the time-honoured irrigation methods, such as benefiting from the gradient of the land or using mechanical methods of lifting and transporting water. But if any such systems existed, they have been obliterated.
A wide range of water structures -- bathing platforms and house drains -- are found in Dholavira, both within and outside the walls of the city. Some are the ordinary kinds also found commonly in several Harappan settlements. But in many Harappan centres, these structures were made of baked bricks because stones were not readily available. Sometimes, the bricks were lined with a cementing clay to render them less porous.
At Dholavira, stones were available in plenty and were used to make tiles for the water structures; drains were made with both ordinary ungeometrical stones as well as neatly cut blocks. There are several channels carefully cut in stone blocks or slabs of varying lengths. Pottery pipes were also used. In some places, mortar was made of superfine, sticky, grey clay. At Mohenjodaro, bitumen, gypsum and lime were used as waterproofing agents.
In the area designated as the Citadel (see box, Walled wonder), an interesting network of drains, both small and large, was discovered. Most of the drains intersect each other and ultimately link up with an arterial drain. The larger ones are high and broad enough to allow a person to walk through them, resembling the drain built to flush out waste water from the Great Bath at Mohenjodaro. The size helped to abate the flow of locked-up water that would otherwise gush out of the bath.
A similar construction was adopted in the 2 principal drains uncovered so far in the Castle portion of the Citadel. One originates from a salient beside the east gate and the other from close to the north gate. Both meet near the centre of the Castle and then run underground towards the west. At the far end, the drain hits a raised pathway that parallels the western wall, then turns south. Near the southwestern corner of the Castle, the drain cuts through the wall and empties out as a cascade into a drain below, which in turn heads further west in the Bailey to a destination that still remains to be exposed.
An interesting feature of the drainage system is the provision of small apertures at short intervals on the roofing slabs of the drain. At first, it was thought that the apertures were meant to allow people to descend and clean the drain, but they were too small. In fact, some of the apertures were far too deep and narrow to be used for regular clean-ups.
There has been some speculation about the purpose of this elaborate network of drains. For one thing, there is no evidence to indicate that they were meant to let out waste water or sewage from the Castle. All the drains, regardless of size, have so far not been found to be linked with residential structures or bathing platforms. Instead, they originate from the fort's walls and streets.
And there are no cesspits in the floor of the drains to trap solid matter. Besides, the Castle is by no means large enough to allow foul-smelling sewage drains to crisscross it. It would have been easier to provide a far more decent and less unpleasant system to get rid of waste.
Another puzzling factor were the apertures. Not only were they to small to let people through, they were also placed too close to each other than would be logically required for cleaning purposes.
If the drains weren't meant to carry out waste water and the apertures not used for cleaning the drains, why was such an intricate network made at all? Excavators from the Archaeological Survey of India (ASI) believe that the drains were designed to collect and carry monsoon run-off to a receptacle for later use. This would also explain the unusual width and height of the drains. The apertures could have acted like ducts to let air escape and facilitate the easy flow of storm water.
Deposits found all along the drains are all freshwater-borne. There was only one exception: near the confluence of the north and east drains, highly decomposed organic matter was found, suggesting its origin in sewage. But here, too, there is a plausible explanation. The organic sediments were recovered along with five human figurines -- mostly of females -- associated with a cult, suggesting that these substances and the figurines were immersed in the water of the drain as part of a ritual.
That apart, there was a time when the drainage system became defunct because of serious damage to its walls as well as the slabs below. An effort was made to revive the system but was soon abandoned. At the time of disuse, some undesirable matter could have accumulated in the drain.
The entire drainage system could have been set up to assiduously conserve every drop of rainwater that fell on the city. Water must have been a treasured substance in an area lacking a perennial source of surface water and where the groundwater, largely brackish and saline in the first place, tends to dry up during droughts.
The inference is also borne out by the fact that the Harappans were extremely conscientious about zealously storing the maximum water -- if not all of it -- that flowed in the flanking streams after the downpours. They created at least 16 reservoirs almost all around within the city walls. A casual estimate indicates about 17 ha, (36 per cent of the walled area) was appropriated for storing water harnessed from streams, by raising several dams -- at least three across the Manhar and two across the Mansar. In addition, there may have been inlet channels penetrating the city walls to bring in surface run-off.
The Manhar and the Mansar have adequate catchment areas, starting from the highest elevation of the northern chain of hills. There are rocky outcrops that were used efficiently for damming the streams. The Harappans reduced the velocity of the water flow further upstream by raising a series of dams, which also diverted surplus water to the reservoirs within the city.
In the whole scheme, the city walls played the most crucial role. First, they served to protect the city. They were strengthened at regular intervals with projecting salients and strong corner towers. There are presumably several gates, which, however, remain to be unearthed.
In addition, the walls -- made of millions of moulded, sum-dried bricks carefully laid in mud-mortar -- acted as effective bunds. At places, the base of the wall is 7 metres thick. Both faces of the wall were plastered with fairly water-repelling, sticky clay. Special and vulnerable areas, mostly on the exterior face, were veneered with hammer-dressed stones.
In order to protect the walls from the water stored in the tanks, a broad road lined with stone-masonry was provided all along. This road served both as a bund and a path for the easy movement of people and materials.
In its heyday, the entire city might have looked like a lake-city or a jala durga (water fort). The area reserved for the tanks was immense, approximately 750 metres long along the southern and northern margins, while the width varied from 70 to 80 metres. In the west, the tank area was about 590 metres long and the width went up to 170 metres in some places. In the southeastern area, for example, the reservoir covered about 5 ha, the largest within the walled area.
A very conservative estimate is that the reservoirs within the city walls covered an area of at least 17 ha, containing not less than 250,000 cubic metres of water. This excludes what might have been stored outside the city.
Keeping in mind the general slope of the city, several bunds were constructed across the width of the tanks to reduce the pressure of the stored water body on the city walls. The bunds also served as causeways for easier movement and in times of scanty rainfall, they enabled water to be stored in selected tanks instead of spreading it out over a large area, which would led to quick evaporation and seepage.
The reservoirs seem to have been desilted regularly and the removed matter was piled up on nearby bunds which, in turn, were protected by masonry walls raised on both sides. Such bunds varied in width from 4.5 metres to 7 metres, depending on their location and use.
To meet exigencies, deep troughs and depressions were cut into the bedrock to store water. At least one such trough was uncovered during test probes. At almost all points, whether in the centre or close to a bund or where inlet channels are presumed to have existed, Harappan pottery was always found stuck to the bedrock; in most places, the bedrock is lower than the present rocky or sandy bed of the flanking runnels. In one of the bunds, near the most conspicuous dam site on the Manhar, a spill-water channel was constructed to carry surplus water into the tanks provided along the south of the Citadel.
Interestingly, a huge artificial reservoir is located near the source of the Mansar, which derived its name from the lake. A massive dam, made of huge stone blocks, that exists there belongs to the early Christian era. A test probe of the lake sediments yielded potsherd of the Christian period and, in addition, two small bits of the late Harappan ones came up from the lowest level. It should not be surprising if the mature Harappans built the original dam, the remains of which may be buried beneath the later construction. On one side of the dam is a rock-cut channel for releasing water to the fields.
From this, it can be surmised that the Harappans stored enough water for daily consumption, construction activities as well as for raising rabi crops. Significantly, M D Kajale, archaeobotanist of the Post-graduate Research Institute, Deccan College, Pune, who is collaborating with the ASI, found carbonised grains of barley and wheat from the strata of all the cultural stages at Dholavira. The land, when given moisture, can produce bumper winter crops, because the sandy loam is rich and fertile.
It seems clear that the Harappan architects were entirely familiar with the basics of hydraulic engineering, which could not have been acquired overnight. The Harappan structures are similar to the gabarbands of south Baluchistan, now attributed to the Magis, the fire-worshipping Zoroastrian priests.
Archaeologically, the gabarbands can be associated with the pre-Harappan settlements because of their geographical proximity. The gabarbands are made of huge boulders and cut-blocks, across gaps in the hills with a broad catchbasin of ephemeral streams. The dams were used to store water or for irrigation. Sometimes, such constructions were raised to conserve soil and moisture.
Environmentally, rocky Baluchistan is similar to the Kutch -- both regions receive short but violent rainstorms, requiring effective water-control measures to ensure rich harvests, especially for winter-cropping. Such techniques may have been inherited by the Harappans of Dholavira, whose ancestry could have originated in Sind and Baluchistan, as attested to by the ceramics that preceded the emergence of classical pottery. There is logic in postulating that the Harappans adopted more or less similar methods in the alluvial plains as well.
R S Bisht is associated with the Indian Institute of Archaeology, New Delhi.