________________
SEPTEMBER, 1932
THE RIVER COURSES OF THE PANJAB AND SIND
169
the order of 5,000 to 500,000 acres. It would be possible even now to travel via Mârot in September after a good monsoon. A traveller or historian would derive very different impressions according to the season, and an explanation based upon a lost river or change in climate might be quite beside the point. In the Panjab the months of May and June are very dry and hot, but rain fell regularly at this period in the year 1917, and the crops sprouted on the threshing floors. This abnormal weather was followed by an exceptional monsoon. At Jagraon in Ludhiana in October 1917, water flowed out of the well heads and the land was too wet to plough; there was excellent recruiting for the Army that autumn. The rivers remained in flood for weeks after the normal time, and the ancient channels and spill-ways in the desert must have been full of water. The travels of a Chinese pilgrim through the Panjab in the Year of Grace 1917 might record an accurate picture of his impressions, but they would be quite misleading.
Another potent factor has been the hand of man. In former days the water of the Ambala streams reached the Sirsa region; their dry channels are still visible, while wells and Persian wheels are found embedded in the sand. A major cause of the change is the deflection of water higher up by dams and inundation canals which checked its course through the Karnal and Patiala levels. Settled conditions and incroased population in recent times have greatly accelerated the spread of grazing and the rate of deforcstation. The condition of the unprotected outer hills has much deteriorated in the last sixty years. Perennial streams have degenerated into sand torrents with a destructive rush of water in the rains, and nothing the rest of the year. There has probably heen more change here in the last hundred years than in the preceding twenty centuries.
It is certain that the course of each of the rivers in the plains of the Panjab and Sind has changed within historical times, but this does not mesu that the main beds have moved to the extent that has been suggested. Harappa is on the old high bank of the Ravi and aerial reconnaissance along the bank has revealed a chain of sites possibly coeval with it. Parappå has turned out to be immemorially older than was suspected when Raverty was riding over the Panjab Bars in 1855, or when Cunningham wrote his Ancient Geography of India in 1871, and it would appear that the old high bank of the Ravi has reniained much as it is today since pre-Vedic times. This great discovery puts the matter in its proper perspective. The bed of a large river in an alluvial plain may be twenty and even thirty miles wide. The river is free to oscillate within these limits, but may not have transgressed them for thousands of years. I think this is true of the Ravi, the Chenab, and the Jhelum.25 The Beas and Sutlej seem to have been more mobile. Major Raverty only had the levels of the Trigonometrical Survey; these were taken here and there, usually at elevated spots, for the purposes of triangulation. They do not give the general slope of the country, and are meaningless for hydraulic purposes. Exact hydraulic data are now to hand. Excavation will determine the real age of the buried towns along the old high banks. I am informed that a mound as far east as Râpar, where the Sutlej leaves the Siwaliks, has been found to belong to the Indus Age. Much new information will be available when the large and fascinating volume envisaged by Dr. Vincent Smith comes to be written. The freaks of even Indian rivers are ultimately governed by levels.
14 There is also Mohenjo-dAro, of course, in the riverain tract of the In.lus. 36 Cp. Cunningham's Ancient Geography of India, p. 223.