The Ancient Geography of India/Multan
|Wikified by:Laxman Burdak, IFS (Retd.)|
4. Multan Province
[p.219]: The southern province of the Panjab is Multan. According to Hwen Thsang it was 4000 li, or 667 miles, in circuit, which is so much greater than the tract actually included between the rivers, that it is almost certain the frontier must have extended beyond them. In the time of Akbar no less than seventeen districts, or separate parganahs, were attached to the province of Multan, of which all those that I can identify, namely, Uch, Dirawal, Moj, and Marot, are to the east of the Satlej. These names are sufficient to show that the eastern frontier of Multan formerly extended beyond the old bed of the Ghagar river, to
[p.220]: the verge of the Bikaner desert. This tract, which now forms the territory of Bahawalpur, is most effectually separated from the richer provinces on the east by the natural barrier of the Great Desert. Under a strong government it has always formed a portion of Multan ; and it was only on the decay of the Muhammadan empire of Delhi that it was made into a separate petty state by Bahawal Khan. I infer, therefore, that in the seventh century the province of Multan must have included the northern half of the present territory of Bahawalpur, in addition to the tract lying between the rivers. The northern frontier has already been defined as extending from Dera Din-panah, on the Indus, to Pak-pattan on the Satlej, a distance of 150 miles. On the west the frontier line of the Indus, down to Khanpur, is 160 miles. On the east, the line from Pak-pattan to the old bed of the Ghagar river, is 80 miles ; and on the south, from Khanpur to the Ghagar, the distance is 220 miles. Altogether, this frontier line is 610 miles. If Hwen Thsang's estimate was based on the short kos of the Panjab, the circuit will be only 21/32 of 667 miles, or 437 miles, in which case the province could not have extended beyond Mithankot on the south.
In describing the geography of Multan it is necessary to bear in mind the great changes that have taken place in the courses of all the large rivers that flow through the province. In the time of Timur and Akbar the junction of the Chenab and Indus took place opposite Uchh, 60 miles above the present confluence at Mithankot. It was unchanged when Rennell wrote his 'Geography of India,' in a.d. 1788, and still later in 1796, when visited by Wilford's surveyor,
[p.221]: Mirza Mogal Beg. But early in the present century the Indus gradually changed its course, and leaving the old channel at 20 miles above Uchh, continued its course to the south-south-west, until it rejoined the old channel at Mithankot.
The present junction of the Ravi River and Chenab takes place near Diwana Sanand, more than 30 miles above Multan ; but in the time of Alexander the confluence of the Hydraotes and Akesines was at a short distance below the capital of the Malli, which I have identified with Multan. The old channel still exists, and is duly entered in the large maps of the Multan division. It leaves the present bed at Sarai Siddhu, and follows a winding course for 30 miles to the south-south-west, when it suddenly turns to the west for 18 miles, as far as Multan, and, after completely encircling the fortress, continues its westerly course for 5 miles below Multan. It then suddenly turns to the south-south- west for 10 miles, and is finally lost in the low-lying lands of the bed of the Chenab. Even to this day the Ravi clings to its ancient channel, and at all high floods the waters of the river still find their way to Multan by the old bed, as I myself have witnessed on two different occasions. The date of the change is unknown; but it was certainly subsequent to the capture of Multan by Muhammed bin Kasim in A.D. 713; and from the very numerous existing remains of canals drawn from the old channel, I infer that the main river must have continued to flow down it within a comparatively recent period, perhaps even as late as the time of Timur. The change, however, must have taken place before the reign of Akbar, as Abul Fazl1
1 ' Ayin Akbari,' ii. 3.
[p.222]: describes the distance from the confluence of the Chenab and Jhelam to that of the Chenab and Ravi as 27 kos, and the distance of the latter from the confluence of the Chenab and Indus as 60 kos, both of which measurements agree with the later state of these rivers.
The present confluence of the Bias and Satlej dates only from about A.D. 1790, when the Satlej finally deserted its old course by Dharmkot, and joined the Bias at Hariki-pattan. For many centuries previously the point of junction had remained constant just above the ferry of Bhao-ki-pattan, between Kasur and Firuzpur. This junction is mentioned by Jauhar in A.D. 1555,1 and by Abul Fazl in 1596.2 But though the confluence of the two rivers near Firuzpur had been long established, yet even at the latter date the waters of the Bias still continued to flow down their old channel, as described by Abul Fazl : — " For the distance of 12 kos near Firuzpur the rivers Biah and Satlej unite, and these again, as they pass on, divide into four streams, the Hur, Hare, Dand, and Nurni, all of which rejoin near the city of Multan." These former beds of the Bias and Satlej still exist, and form a most complicated network of dry channels, covering the whole of the Doab between the Satlej and the high bank of the old Bias. None of the names given in Gladwyn's translation of the ' Ayin Akbari' are now to be found ; but I am inclined to attribute this solely to the imperfection of the Persian alphabet, which is a constant source of error in the reading of proper names. The Har I would identify with the Par, the Hari with the Raghi, and the Nurni with
1 ' Memoirs of Humayun,' p. 113. 2 ' Ayin Akbari,' ii. 108.
[p.223]: the Suk-Nai, all dry beds of the Bias river to the south of Harapa. The Dand is probably the Dhamak, or Dank, an old channel of the Satlej, which in its lower course takes the name of Bhatiyari, and passing by Mailsi, Kahror, and Lodhran, joins the present channel just above its confluence with the Chenab. In most of our maps the Old Bias is conducted into the lower course of the Bhatiyari, whereas its still existing and well-defined channel joins the Chenab 20 miles below Shujahabad, and its most southerly point is 10 miles distant from the nearest bend of the Bhatiyari.
The changes just described are only the most prominent fluctuations of the Panjab rivers, which are constantly shifting their channels. The change in the Bias is the most striking, as that river has altogether lost its independent course, and is now a mere tributary of the Satlej. But the fluctuations of the other rivers have been very remarkable. Thus, the valley of the Chenab below Kalowal is nearly 30 miles broad, and that of the Ravi, near Gugera, is 20 miles, the extreme limits of both rivers being marked by well-defined high banks, on which are situated many of the most ancient cities of the Panjab.
In the Multan division these old sites are very numerous, but they are now mostly deserted and nameless, and were probably abandoned by their inhabitants as the rivers receded from them. This was certainly the case with the old town of Tulamba, which is said to have been deserted so late as 150 years ago, in consequence of a change in the course of the Ravi, by which the water supply of the town was entirely cut off. The same cause, but at a much earlier date, led to the
[p.224]: desertion of Atari, a ruined town 20 miles to the west- south-west of Tulamba, which was supplied by a canal from the old bed of the Ravi. The only places which I think it necessary to notice in the present account are the following : —
|Bari Doab||1. Tulamba.|
|Jalandhar Pith||4. Kahror.|
|At junction||5. Uchh.|
Four of these places are celebrated in the history of India, and the second, named Atari, I have added on account of its size and position, which would certainly have attracted the notice of Alexander and other conquerors of the Panjab.
[p.224]: The town of Tulamba is situated on the left bank of the Ravi, at 52 miles to the north-east of Multan. It is surrounded with a brick wall, and the houses are built chiefly of burnt bricks, brought from the old fort of Tulamba, which is situated one mile to the south of the present town. According to Masson,1 this "must have been in the ancient time a remarkably strong fortress," which it undoubtedly was, as Timur left it untouched, because its siege would have delayed his progress.2 It is curious that it escaped the notice of Burnes, as its lofty walls, which can be seen from a great distance, generally attract the attention of travellers. I have visited the place twice. It consisted of an open city, protected on the south by
1 ' Travels,' i. 456. 2 Briggs's ' Ferishta,' i. 487.
[p.225]: a lofty fortress 1000 feet square. The outer rampart is of earth, 200 feet thick, and 20 feet high on the outer face, or faussebraie, with a second rampart of the same height on the top of it. Both of these were originally faced with large bricks, 12 by 8 by 2½ inches. Inside the rampart there is a clear space, or ditch, 100 feet in breadth, surrounding an inner fort 400 feet square, with walls 40 feet in height, and in the middle of this there is a square tower or castle, 70 feet in height, which commands the whole place. The numerous fragments of bricks lying about, and the still existing marks of the courses of bricks in many places on the outer faces of the ramparts, confirm the statements of the people that the walls were formerly faced with brick. I have already mentioned that this old fort is said to have been abandoned by the inhabitants about 300 years ago, in consequence of the change in the course of the Ravi, which entirely cut off their supply of water. The removal is attributed to Shujawal Khan, who was the son-in-law and minister of Mahmud Langa of Multan, and the brother-in-law of his successor, from about A.D. 1510 to A.D. 1525.
The antiquity of Tulamba is vouched for by tradition, and by the large size of the bricks, which are similar to the oldest in the walls and ruins of Multan. The old town was plundered and burnt by Timur, and its inhabitants massacred ; but the fortress escaped his fury, partly owing to its own strength and partly to the invader's impatience to continue his march towards Delhi. There is a tradition that Tulamba was taken by Mahmud of Ghazni, which is very probably true, as it would have been only a few miles out of his
[p.226]: direct route to Multan. For the same reason I am led to believe that it must have been one of the cities captured by Alexander. Masson1 has already suggested that it represents "the capital of the Malli," or perhaps "the fort held by Brahmans, whose defence was so obstinate and so fatal to themselves, and which was evidently contiguous to the capital of the Malli." But as I do not agree with either of these suggestions, I will now examine and compare the different accounts of this part of Alexander's route.
In my account of Kot Kamalia I adduced some strong reasons for identifying that place with the first city captured by Alexander on his march from the junction of the Hydaspes and Akesines against the Malli. Arrian then relates that " Alexander, having allowed his soldiers some time for refreshment and rest, about the first watch of the night set forward, and marching hard all that night came to the river Hydraotes about daylight, and understanding that some parties of the Malii were just passing the river, he immediately attacked them and slew many, and having passed the river himself with his forces in pursuit of those who had gained the further side, he killed vast numbers of them and took many prisoners. However, some of them escaped, and betook themselves to a certain town well fortified both by art and nature."
A whole night's march of eight or nine hours could not have been less than twenty-five miles, which is the exact distance of the Ravi opposite Tulamba from Kot Kamalia. Here then I infer that Alexander must have crossed the Ravi ; and I would identify Tulamba itself with the " town well fortified both by art and
1 ' Trayels,' i. 456. 2 ' Anabasis,' vi. 3.
[p.227]:nature," the art being the brick walls, and the nature, the enormous mounds of earthen ramparts. The account of Curtius1 agrees with that of Arrian, " on the bank of a river another nation mustering forty thousand infantry opposed him. Crossing the river he put them to flight, and stormed the fort in which they took refuge."
Diodorus relates the same story of a people named Agalassae, who opposed Alexander with forty thousand infantry and three thousand cavalry. All these accounts evidently refer to the same place, which was a strong fort near the left bank of the Ravi. This description would apply also to Harapa ; but I have already shown that Harapa was most probably the city against which Perdikkas was detached ; besides which it is not more than 16 miles distant from Kot Kamalia. Tulamba, on the contrary, fulfills all the conditions ; and is also on the high-road to Multan, the capital of the Malli, against which Alexander was then proceeding.
The name of Agalassae or Agalessensae is puzzling. According to Arrian the people of the town were the Malli, but it may be remarked that neither the Oxudrakae nor the Malli are mentioned by Diodorus and Curtius until later. Justin couples a people called Gesteani with the Arestae or Kathaei, who should therefore be the same as the Malli or Oxudrakae, but they are not mentioned by any other author. Agala or Agalassa might be the name of the town itself, but unfortunately it has no similarity with Tulamba, or with any other place in the neighbourhood.
1 Vita Alex., ix. 4, 10. The text has in ripa fluminum, which is an obvious mistake for fluminis, as is proved by the use of amne immediately following.
[p.228]: The third city captured by Alexander in his campaign against the Malli is described in similar terms by all the historians. According to Arrian1 " Alexander then led his army against a certain city of the Brachmani, where he heard another body of the Malli had fled." The garrison " abandoned the city and fled to the castle," which being stormed they set fire to their houses, and perished in the flames. "About 5000 of them fell during the siege, and so great was their valour that few came alive into the enemy's hands." Both Curtius2 and Diodorus3 mention the fire, and the stout defence made by the garrison, which the latter author numbers at 20,000 men, of whom 3000 only escaped by taking refuge in the citadel, where they capitulated. Curtius also states that the citadel was uninjured, and that Alexander left a garrison in it.4
All these accounts agree very well with the position and size of the old ruined town and fort of Atari, which is situated 20 miles to the west-south-west of Tulamba, and on the high-road to Multan. The remains consist of a strong citadel 750 feet square and 35 feet high, with a ditch all round it, and a tower in the centre 50 feet high. On two sides are the remains of the town forming a mound 20 feet high, and 1200 feet square, the whole being a mass of ruin 1800 feet m length, and 1200 feet in breadth. Of its history there is not even a tradition, but the large size of the bricks
1 'Anabasis,' vi. 7.
2 Vita Alex., ix. 4, 10.
3 Hist., xvii. 52.
4 Vita Alex , ix. 4. " Arx erat oppidi intacta, ia qua praesidium dereliquit.'
[p.229]: is sufficient to show that it must be a place of considerable antiquity. The name of the old city is quite unknown. Atari is simply that of the adjacent village, which is of recent origin, having been established by a member of the Atariwala family of Sikhs. But judging from its size and strength, and its very favourable position between Tulamb and Multan, I think that the ruined mound of Atari has a very good claim to be identified with the strong city of the Brahmans which made so stout a defence against Alexander.
Curtius adds some particulars about this city, which are not even alluded to by either Arrian or Diodorus ; but they are still deserving of consideration, as they may perhaps be founded on the statements of one of the companions. He states that Alexander " went completely round the citadel in a boat," which is probable enough, as its ditch was no doubt capable of being filled at pleasure with water from the Ravi, as was actually the case with the ditch of Multan. Now the old citadel of Atari is still surrounded by a ditch which could easily have been filled from some one of the old canals that pass close by the place. The number of these canal beds is most remarkable ; I counted no less than twelve of them in close parallel lines immediately to the west of Atari, all of them drawn from the old bed of the Ravi to the south of Sarai Siddhu. I am therefore quite prepared to admit the probability that the city of the Brahmans was surrounded by a wet ditch on which Alexander embarked to inspect the fortifications. But when Curtius adds that the three greatest rivers in India, except the Ganges, namely the Indus, the Hydaspes, and the Akesines, joined their waters to
[p.230]: form a ditch round the castle,1 I can only suppose either that the passage has been accidentally transferred from the account of some later siege of a city situated below the confluence of the Five Rivers, or that the author has mixed up into one account two and perfectly distinct statements concerning the ditches of the fort and the confluence of the rivers. Diodorus also describes the junction of the rivers, but as he makes no allusion to their waters forming a ditch about the fort, it is quite possible that this account of three rivers may be due to the inflated imagination of Curtius.
[p.230]: The famous metropolis of Multan was originally situated on two islands in the Ravi, but the river has long ago deserted its old channel, and its nearest point is now more than 30 miles distant. But during high floods the waters of the Ravi still flow down their old bed, and I have twice seen the ditches of Multan filled by the natural overflow of the river.2 Multan consists of a walled city and a strong fortress, situated on opposite banks of an old bed of the Ravi, which once flowed between them as well as around them. The original site consisted of two low mounds not more
1 Vila Alex., ix. 4. " Ipse navigio circumvectus est areem; quippe tria flumina, tota India praeter Gangen maxima, munimento arcis applicant undas. A septentrione Indus alluit ; a meridie Acesines Hydaspi confunditur."
2 Burnes, 'Travels in the Punjab, Bokhara,' etc. i. 97, erroneously attributes the inundation of the country around Multan to the " Chenab and its canals." If he had travelled by land instead of by the river, he would have seen that the inundation is due to the flood waters of the Ravi resuming their ancient course from Sarai Siddhu direct upon Multan. I travelled over this line in the end of August, 1856, and saw the old bed of the Ravi in full flood.
[p.231]: than 8 or 10 feet high above the general level of the country. The present height varies from 45 to 50 feet, the difference of 35 to 40 feet being simply the accumulation of rubbish during the lapse of many centuries. This fact I ascertained personally by sinking several wells down to the level of the natural soil, that is, of soil unmixed with bricks, ashes, and other evidences of man's occupation.
The citadel may be described as an irregular semicircle, with a diameter, or straight side of 2500 feet facing the north-west, and a curved front of 4100 feet towards the city, making a circuit of 6600 feet, or just one mile and a quarter. It had 46 towers or bastions, including the two flanking towers at each of the four gates. The walled city, which envelopes the citadel for more than two-thirds of the curve, is 4200 feet in length, and 2400 feet in breadth, with the long straight side facing the south-west. Altogether the walled circuit of Multan, including both city and citadel, is 15,000 feet, or very nearly 3 miles, and the whole circuit of the place, including its suburbs, is 4½ miles. This last measurement agrees very nearly with the estimate of Hwen Thsang, who makes the circuit of Multan 30 li, or just 5 miles.1 It agrees even more exactly with the estimate of Elphinstone, who, with his usual accuracy, describes Multan as "above four-miles and a half in circumference,"2 The fortress had no ditch when seen by Elphinstone and Burnes, as it was originally surrounded by the waters of the Ravi. But shortly after Burnes's visit, a ditch was added by Sawan Mall, the energetic governor of Ranjit Singh. The walls are said to have been built by Murad Baksh,
1 Julien's ' Hiouen Thsang,' iii. 173. 2 ' Kabul," i. 27.
[p.232]: the youngest son of Shah Jahan ; but when I dismantled the defences of Multan in 1854, I found that the walls were generally double, the outer wall being about 4 feet thick, and the inner wall 3½ feet to 4 feet.1 I conclude, therefore, that only the outer wall, or facing, was the work of Murad Baksh. The whole was built of burnt bricks and mud, excepting the outer courses, which were laid in lime-mortar to a depth of 9 inches. Multan is known by several different names, but all of them refer either to Vishnu or to the Sun, the latter being the great object of worship in the famous temple that once crowned the citadel. Abu Rihan mentions the names of Kasyapa-puru, Hansapura, Bhagapura, and Sambapura, to which I may add, Prahladapura and Adyasthana.
According to the traditions of the people, Kasyapa-pura was founded by Kasyapa, who was the father of the twelve Adityas, or Sun-gods, by Aditi, and of the Daityas, or Titans, by Diti. He was succeeded by his eldest son, the Daitya, named Hiranya-Kasipu, who is famous throughout India for his denial of the omnipresence of Vishnu, which led to the manifestation of the Nara-Sinha, or "Man-lion" avatar. He was followed by his still more famous son Prahlada, the ardent worshipper of Vishnu, after whom the city was named Prahladapura. His great-grandson, Bana, commonly called Bana the Asur (Banasura), was the unsuccessful antagonist of Krishna, who took possession of the kingdom of
1 It may be interesting to note that on dismantling the wall near the Sikhi Darwaza, or " Spiked Gate," I found the only two shot that were fired from the great one hundred-pounder gun, which the Bhangi Misal of of Sikhs brought against Multan in the beginning of this century. The two shot had completely penetrated through the brick wall of 7 feet, and were within three feet of each other.
[p.233]: Multan. Here Samba, the son of Krishna, established himself in the grove of Mitra-vana, and by assiduous devotion to Mitra, or the " Sun," was cured of his leprosy. He then erected a golden statue of Mitra, in a temple named Adyasthana (आद्यस्थान), or the " First Shrine," and the worship of the Sun thus began by Samba, has continued at Multan down to the present day.
The story of Samba, the son of Krishna, is told in the Bhavishya Purana1 but as it places the Mitra-vana, or "Sun-grove," on the bank of the Chandrabhaga, or Chenab river, its composition must be assigned to a comparatively late period, when all remembrance of the old course of the Ravi flowing past Multan had died away. We know, however, from other sources, that the Sun-worship at Multan must be very ancient.
In the seventh century Hwen Thsang found a magnificent temple with a golden statue of the god most richly adorned, to which the kings of all parts of India sent offerings. Hence the place became commonly known amongst the early Arab conquerors as " The Golden Temple ;" and Masudi even affirms that el Multan means " meadows of gold."2 Hwen Thsang calls it Meu-lo-san-pu-lo, which, according to M. Vivien de St. Martin, is a transcription of Mulasthanipura. The people themselves refer the name to Mula-sthana, which agrees with the form of Mula-lana, quoted by Abu Rihan from a Kashmirian writer. Mūla means
1 Wilford, ' Asiatic Researohes,' xi. 69 ; and H. H. "Wilson, in Reinaud, ' Memoire sur l'Inde,' p. 392.
2 Masudi, ' Gildemeister,' p. 134: " domum auream : " so also Sir H. M. Elliot, ' Muhammadan Historians," p. 56 ; but at p. 57 he translates "golden temple." Prof. Dowson, i. 33, has "boundary of the house of gold," translating Masudi ; and at i. 81, " the house of gold," translating Idrisi.
[p.234]: "root, or origin," and sthana, or than, in the spoken dialects, means " place, or shrine." Hence, Mulasthana is the "Temple of Mula,'" which I take to be an appellation of the Sun. In the Amarakosha one of the names of the Sun is Vradhna, which is also given as a synonym of Mula ; hence vradhna must be connected with the Latin radix and radius, and also with the Greek ράβδος . But as radix signifies not only origin, or root, in general, but also a particular root, the radish, so also does mula signify origin, or root, and mulaka, or muli, a radish. The connection between a sunbeam and a radish obviously lies in their similarity of shape, and hence the terms radius and mula are both applied to the spoke of a wheel. Mula-sthana is said by Wilson to mean "heaven, ether, space, atmosphere, God," any one of which names would be applicable to the Sun as the lord of the ethereal space. For these reasons I infer that mula is only an epithet of the Sun, as the God of rays, and that Mula-sthana-pura means simply the "city of the Temple of the Sun.
The earliest name is said to have been Kasyapapura, or as it is usually pronounced, Kasappur, which I take to be the Kaspa-puros of Hekataeus, and the Kaspaturos of Herodotus, as well as the Kaspeira of Ptolemy. The last town is placed at a bend on the lower course of the Rhuadis, or Ravi, just above its junction with the Sandobag, or Chandrabhaga. The position of Kaspeira therefore agrees most exactly with that of Kasyapapura or Multan, which is situated on the old bank of the Ravi, just at the point where the channel changes its course
[p.235]: from south-east to east. This identification is most important, as it establishes the fact that Multan or Kaspeira, in the territory of the Kaspeirei, whose dominion extended from Kashmir to Mathura, must have been the principal city in the Panjab towards the middle of the second century of the Christian era. But in the seventh century it had already acquired the name of Mulasthanapura, or Multan, which was the only name known to the Arab authors down to the time of Abu Rihan, whose acquirement of Sanskrit gave him access to the native literature, from which he drew some of the other names already quoted. The name of Adyasthana, or " First Shrine," is applied in the Bhavishya Purana to the original temple of the Sun, which is said to have been built by Samba, the son of Krishna; but adya is perhaps only a corruption of Aditya, or the Sun, which is usually shortened to adit, and even ait, as in aditwar and aitwar for Adityawara, or Sunday. Biladuri calls the idol a representation of the prophet Job, or Ayub, which is an easy misreading of for adit. Prahladapura, or Pahladpur, refers to the temple of the Narsingh Avatar, which is still called Pahladpuri. When Burnes was at Multan, this temple was the principal shrine in the place, but the roof was thrown down by the explosion of the powder magazine during the siege in January, 1849, and it has not since been repaired. It stands at the north-eastern angle of the citadel, close to the tomb of Bahawal Hak. The great temple of the Sun stood in the very middle of the citadel, but it was destroyed during the reign of Aurangzib, and the Jamai Masjid was erected on its site. This masjid was the powder magazine of the Sikhs, which was blown up in 1849.
[p.236]: By the identification of Kasyapapura with, the Kaspeira of Ptolemy I have shown that Multan was situated on the bank of the Ravi in the first half of the second century of the Christian era. Hwen Thsang unfortunately makes no mention of the river ; but a few years after his visit the Brahman Rajah of Sindh, named Chach, invaded and captured Multan, and the details of his campaign show that the Ravi still continued to flow under its walls in the middle of the seventh century. They show also that the Bias then flowed in an independent channel to the east and south of Multan. According to the native chronicles of Sindh, Chach advanced to Pabiya, or Bahiya1 on the south bank of the Bias, from whence he advanced to Sukah or Sikkah on the bank of the Ravi, at a short distance to the eastward of Multan. This place was soon deserted by its defenders, who retired towards Multan, and joined Raja Bajhra in opposing Chach on the banks of the Ravi. After a stout fight the Multanis were defeated by Chach, and retired into their fortress, which after a long siege surrendered on terms.2
This brief notice of the campaign of Chach will now enable us to understand more clearly the campaign of Alexander against the capital of the Malli. My last notice left him at the strong Brahman city, which I have identified with Atari, 34 miles to the north-east of Multan, and on the high-road from Tulamba. Here I will resume the narrative of Arrian.3
2 Lieut. Postans in Journ. Asiat. Soc. Bengal, 1838, p. 94. Sir H. W. Elliot, ' History of India,' edited by Prof. Dowson, i. 143.
3 ' Anabasis,' vi. 8.
- [p.237]: " Having tarried there one day to refresh his army, he then directed his march against others of the same nation, who, he was informed, had abandoned their cities and retired into the deserts ; and taking another day's rest, on the next he commanded Python, and Demetrius the captain of a troop of horse, with the forces they then had, and a party of light armed foot, to return immediately to the river, etc. In the meanwhile he led his forces to the capital city of the Malli, whither, he was informed, many of the inhabitants of other cities had fled for their better security."
Here we see that Alexander made just two marches from the Brahman city to the capital, which agrees very well with the distance of 34 miles between Atari and Multan. In searching for the chief city of the Malli or Malii, we must remember that Multan has always been the capital of the Lower Panjab, that it is four times the size of any other place, and is undeniably the strongest fort in this part of the country. All these properties belonged also to the chief city of the Malli. It was the capital of the country ; it had the greatest number of defenders, 50,000 according to Arrian, and was therefore the largest place ; and lastly, it must have been the strongest place, as Arrian relates that the inhabitants of other cities had fled to it "for better security." For these reasons I am quite satisfied that the capital city of the Malli was the modern Multan; but the identification will be still further confirmed as we proceed with Arrian's narrative.
On Alexander's approach the Indians came out of their city, and " crossing the river Hydraotes, drew up their forces upon the bank thereof, which was steep and difficult of ascent, as though they would have
[p.238]: obstructed his passage . . . when he arrived there, and saw the enemy's army posted on the opposite bank, he made no delay, but instantly entered the river with the troops of horse he had brought with him." The Indians at first retired ; " but when they perceived that their pursuers were only a party of horse, they faced about and resolved to give him battle, being about 50,000 in number." From this account I infer that Alexander must have advanced upon Multan from the east, his march, like that of Chach, being determined by the natural features of the country. Now the course of the old bed of the Ravi for 18 miles above Multan is almost due west, and consequently Alexander's march must have brought him to the fort of Sukah or Sikkah, which was on the bank of the Ravi at a short distance to the east of Multan. From this point the same narrative will describe the progress of both conquerors. The town on the east bank of the Ravi was deserted by its garrison, who retired across the river, where they halted and fought, and being beaten took refuge in the citadel. The fort of Sukah must have been somewhat near the present Māri Sital, which is on the bank of the old bed of the Ravi, 2½ miles to the east of Multan.
Alexander was dangerously wounded by Mallis - At the assault of the capital Alexander was dangerously wounded, and his enraged troops spared neither the aged, nor the women, nor the children, and every soul was put to the sword. Diodorus and Curtius assign this city to the Oxudrakae ; but Arrian distinctly refutes this opinion,1 "for the city," he says " belonged to the Malli, and from that people he received the wound. The Malli indeed designed to
1 Anabasis,' vi. 11.
[p.239]: have joined their forces with the Oxudrakae, and so to have given him battle ; but Alexander's hasty and unexpected march through the dry and barren waste prevented their union, so that they could not give any assistance to each other." Strabo also says that Alexander received his wound at the capture of a city of the Malli.1
When Alexander opened his campaign against the Malli, he dispatched Hephaestion with the main body of the army five days in advance, with orders to await his arrival at the confluence of the Akesines and Hydraotes.2 Accordingly after the capture of the Mallian capital, " as soon as his health would admit, he ordered himself to be conveyed to the banks of the river Hydraotes, and from thence down the stream to the camp, which was near the confluence of the Hydraotes and Akesines, where Hephaestion had the command of the army and Nearchus of the navy." Here he received the ambassadors from the Oxudrakae and Malli tendering their allegiance. He then sailed down the Akesines to its confluence with the Indus, where he " tarried with his fleet till Perdikkas arrived with the army under his command, having subdued the Abastani, one of the free nations of India, on his way."
At the capture of Multan by Chach, in the middle of the seventh century, the waters of the Ravi were still flowing under the walls of the fortress, but in A.D. 713, when the citadel was besieged by Muhammad bin Kasim, it is stated by Biladuri 3 that " the city was supplied with water by a stream flowing from the
1 Geogr., xv. 1, 33.
2 ' Anabasis,' vi. 5.
3 Reinaud, ' Fragments Arabes,' p. 199.
[p.240]: river (name left blank by M. Reinaud) ; Muhammad cut off the water, and the inhabitants, pressed by thirst, surrendered at discretion. All the men capable of bearing arms were put to death, and the women and children, with 6000 priests of the temple, were made slaves." The canal is said to have been shown to Muhammad by a traitor. I am willing to accept this account as a proof that the main stream of the Ravi had already deserted its old channel ; but it is quite impossible that Multan could have been forced to surrender from want of water. I have already explained that one branch of the Ravi formerly flowed between the city and fortress of Multan, and that the old bed still exists as a deep hollow, in which water can be reached at most times by merely scratching the surface, and at all times by a few minutes' easy digging. Even in the time of Edrisi1 the environs of the town are said to have been watered by a small river, and I conclude that some branch of the Ravi must still have flowed down to Multan. But though the narrative of Biladuri is undoubtedly erroneous as to the immediate cause of surrender, I am yet inclined to believe that all the other circumstances may be quite true. Thus, when the main stream of the Ravi deserted Multan, the city, which is still unwalled on the side towards the citadel, must have been protected by continuing its defences right across the old bed of the river to connect them with those of the fortress. In these new walls, openings must have been left for the passage of the waters of the canal or branch of the Ravi, whichever it may have been, similar to those which existed in modern times. Edrisi specially notes
1 Geogr., Jauberts trauslation, i. 168.
[p.241]: that Multan was commanded by a citadel, which had four gates, and was surrounded by a ditch. I infer, therefore, that Muhammad Kasim may have captured Multan in the same way that Cyrus captured Babylon, by the diversion of the waters which flowed through the city into another channel. In this way he could have entered the city by the dry bed of the river, after which it is quite possible that the garrison of the citadel may have been forced to surrender from want of water. At the present day there are several wells in the fortress, but only one of them is said to be ancient ; and one well would be quite insufficient for the supply even of a small garrison of 5000 men.
[p.241]: The ancient town of Kahror is situated on the southern bank of the old Bias river, 50 miles to the south-east of Multan, and 20 miles to the north-east of Bahawalpur. It is mentioned as one of the towns which submitted to Chach1 after the capture of Multan in the middle of the seventh century. But the interest attached to Kahror rests on its fame as the scene of the great battle between Vikramaditya and the Sakas, in A.D. 79. Abu Rihan describes its position as situated between Multan and the castle of Loni. The latter place is most probably intended for Ludhan, an ancient town situated near the old bed of the Satlej river, 44 miles to the east-north-east of Kahror, and 70 miles to the east-south-east of Multan. Its position is therefore very nearly halfway between Multan and Ludhan, as described by Abu Rihan.
[p.242]:The old town of Uchh is situated on the eastern bank of the Panjnad, 70 miles to the south-south-west of Multan, and 45 miles to the north-east of the present confluence with the Indus at Mithunkot. The change in the course of the Indus has taken place since the time of "Wilford's surveyor, Mirza Mogal Beg, who surveyed the Panjab and Kabul between the years 1786 and 1796, and this part in 1787-88. The former channel still exists under the name of Nala Puran, or the " Old Stream." Uchcha means " high, lofty," both in Sanskrit and in Hindi ; and Uchchangar is therefore a common name for any place situated on a height. Thus we have Uchchagaon or Bulandshahr, as the Muhammadans call it, on the high bank of the Kali Nadi, 40 miles to the south-east of Delhi. "We have another Uchh on a mound to the west of the confluence of the Chenab and Jhelam ; and a third Uchh, which is also situated on a mound, is the subject of the present description. According to Burnes, 1 Uchh is formed of three distinct towns, a few hundred yards apart from each other, and each encompassed by a brick wall, now in ruins. Masson2 mentions only two separate towns ; but the people themselves say that there were once seven different towns named Uchchnnagar. In Mogal Beg's map Uchh is entered with the remark, " consisting of seven distinct villages." According to Masson, Uchh is chiefly "distinguished by the ruins of the former towns, which are very extensive, and attest the pristine prosperity of the locality." According to Burnes,
1 ' Bokhara,' i. 79. 2 ' Travels,' i. 22,
[p.243]: the town of Uchh stands on a mound, which he judged, from a section exposed by an inundation of the Chenab, to be formed of the ruins of houses. This opinion is doubtless correct, as the place has been repeatedly destroyed and rebuilt. After the last great siege, in A.H. 931, or A.D. 1524-25, by Husen Shah Arghun, the walls of Uchh were levelled to the ground, and the gates and other materials were carried off to Bakar in boats. 1 Its favourable position at the old confluence of the Panjab rivers must have made it a place of importance from the earliest times. Accordingly, we learn from Arrian that Alexander " ordered a city to be built at the confluence of the two rivers, imagining that by the advantage of such a situation it would become rich and populous."2 It is probably this city which is mentioned by Rashid ud din, as the capital of one of the four principalities of Sindh under Ayand, the son of Kafand, who reigned after Alexander. He calls the place Askaland-usah, which would be an easy corruption of Alexandria Uchcha, or Ussa, as the Greeks must have written it. I think, also, that Uchh must be the Iskandar, or Alexandria, of the Chach-namah, which was captured by Chach on his expedition against Multan. 3 After the Muhammadan conquest the place is mentioned only by its native name of Uchh. It was captured by Mahmud of Ghazni, and Muhammad Ghori, and it was the chief city of Upper Sindh under Naser ud din Kubachah. At a later period it formed part of the independent kingdom of Multan, which was established shortly
1 Postans, Journ. Asiat. Soc. Bengal, 1841, p. 275.
2 ' Anabasis,' vi. 15.
3 Postans, Journ. Asiat. Soc. Bengal, 1838, p. 94.
[p.244]: after the troubles that followed the invasion of Timur. 1 In A.D. 1524 it was taken by storm by Shah Husen or Hasan Arghun of Sindh, when its walls were dismantled, as I have already noticed. But after the capture of Multan, Husen ordered the fort of Uchh to be rebuilt, in which he left a large garrison to secure the possession of his recent conquests. In the reign of Akbar, Uchh was permanently annexed to the Mogal empire, and is included by Abul Fazl amongst the separate districts of the Subah of Multan.
The country at the confluence of the Panjab rivers is assigned by Curtius to the Sambracae or Sabracae, and by Diodorus to the Sambastae. They are not mentioned by Arrian, at least under this name ; but I think that the Ossadii, who tendered their allegiance to Alexander at the confluence of the rivers, were the same people. It is probable also that the Abastani, who were subdued by Perdikkas, belonged to the same class. Perdikkas had been dispatched by Alexander to the east of the Ravi, where he captured a town which I have identified with Harapa. I infer that his campaign must have been an extended one, as Alexander, whose own movements had been delayed by his wound, was still obliged to halt for him at the confluence of the rivers. It seems highly probable therefore that he may have carried the Greek arms to Ajudhan on the banks of the Satlej, from which his march would have been along the course of that river by Ludhan, Mailsi, Kahror, and Lodhran, to Alexander's camp at Uchh. In this route he must have encountered the Johiya Rajputs, who have occupied
1 Briggs's ' Ferishta,' iv, 380.
[p.245]: both banks of the Satlej from Ajudhan to Uchh from time immemorial. I think therefore that the Abastani, whom Perdikkas subdued have a strong claim to be identified with the Johiya Rajputs. The country about Multan is still called Johiya-bar or Yaudheya-wara.
The Johiyas are divided into three tribes, named Langavira or Lakvira, Madhovira or Madhera, and Adamvira or Admera. The Sambracae would appear to have been divided into three clans, as being a free people without kings they chose three generals to lead them against the Greeks. Now Johiya is an abbreviation of Jodhiya, which is the Sanskrit Yaudheya, and there are coins of this clan of as early a date as the first century of the Christian era, which show that the Yaudheyas were even then divided into three tribes. These coins are of three classes, of which the first bears the simple inscription Jaya-Yaudheya-ganasya, that is (money) " of the victorious Yaudheya tribe. The second class has dwi at the end of the legend, and the third has tri, which I take to be contractions for dwitiyasya and tritiyasya, or second and third, as the money of the second and third tribes of the Yaudheyas. As the coins are found to the west of the Satlej, in Depalpur, Satgarha, Ajudhan, Kahror, and Multan, and to the eastward in Bhatner, Abohar, Sirsa, Hansi, Panipat, and Sonipat, it is almost certain that they belong to the Johiyas, who now occupy the line of the Satlej, and who were still to be found in Sirsa as late as the time of Akbar. The Yaudheyas are mentioned in the Allahabad inscription of Samudra Gupta, and at a still earlier date by Panini in the Junagarh inscription of Rudra Dama. 1 Now the great grammarian was
1 Dr. Bhau Daji in ' Bombay Journal,' vii. 120.
[p.246]: certainly anterior to Chandra Gupta Maurya, and his mention of the Yaudheyas proves that they must have been a recognised clan before the time of Alexander. The inscription of Rudra Dama, in which he boasts of having "rooted out the Yaudheyas, shows that this powerful clan must have extended their arms very far to the south, otherwise they would not have come into collision with the princes of Surashtra. From these facts I am led to infer that the possessions of the Johiyas in the time of Alexander most probably extended from Bhatner and Pakpatan to Sabzalkot, about halfway between Uchh and Bhakar.
I will now examine the different names of the people who made their submission to Alexander during his halt at the confluence of the Panjab rivers. According to Curtius they were called Sambracae or Sabracae ; 1 according to Orosius Sabagrae ; and according to Diodorus, who placed them to the east of the river, Sambastae. 2 They were a powerful nation, second to none in India for courage and numbers. Their forces consisted of 60,000 foot, 6000 horse, and 500 chariots. The military reputation of the clan suggests the probability that the Greek name may be descriptive of their warlike character, just as Yaudheya means " warrior or soldier." I think, therefore, that the true Greek name may have been Sambagrae, for the Sanskrit Samvagri, that is, the "united warriors," or <greek> which, as they were formed of three allied tribes, would have been an appropriate appellation. In confirmation of this suggestion, I may note the fact that
1 Vita Alex., ix. 8. "Inde Sabracas adiit, Talidam Indiae gentem, qua' populi, non regum, imperio regebatur." 2 Hist., xvii. 10.
the country of which Bikaner is now the capital was originally called Bagar-des, or the land of the Bagri, or " Warriors," whose leader was Bagri Rao. 1 Bhati also means " warrior or soldier." We thus find three tribes at the present day, all calling themselves " warriors," who form a large proportion of the population in the countries to the east of the Satlej ; namely, Johiyas or Yaudheyas along the river, Bagris in Bikaner, and Bhatis in Jesalmer. All three are of acknowledged Lunar descent ; and if my suggested interpretation of Sambagri be correct, it is possible that the name might have been applied to these three clans, and not to the three tribes of the Yaudheyas. I think, how-ever, that the Yaudheyas have a superior claim, both on account of their position along the banks of the Satlej, and of their undoubted antiquity. To them I would attribute the foundation of the town of Ajudhan, or Ayodhanam, the " battle-field," which is evidently connected with their own name of Yaudheya, or Ajudhiya, the " warriors." The latter form of the name is most probably preserved in the Ossadii of Arrian, a free people, who tendered their allegiance to Alexander at the confluence of the Panjab rivers. The Ossadii of Arrian would therefore correspond with the Sambastae of Diodorus and the Sambracae of Curtius, who made their submission to Alexander at the same place. Now Ossadioi or Assodioi is as close a rendering of Ajudhiya as could be made in Greek characters. We have thus a double correspondence both of name and
1 This information I obtained at the famous fortress of Bhatner in the Bikaner territory. The name is certainly as old as the time of Jahangir, as Chaplain Terry describes ' Bikaneer ' as the chief city of ' Bakar.' See 'A Voyage to East India,' p. 86.