The Ancient Geography of India/Taki
|Wikified by:Laxman Burdak, IFS (Retd.)|
The kingdom of Taki
[p.148]: The kingdom which Hwen Thsang calls Tse-kia , or Taki (टाकी), embraced the whole of the plains of the Panjab from the Indus to the Bias, and from the foot of the mountains to the junction of the five rivers below Multan.1 The Chinese syllable tse is used by Hwen Thsang to represent the cerebral ṭ of the Sanskrit in the name of Danakakaṭa which is found in no less than five of the western cave inscriptions at Kanhari and Karli.2 In Hwen Thsang's travels this name is written To.no. kia-tse.kia, in which the last two syllables are transposed. It is the Danaka of Abu Rihan, which, as will be shown hereafter, is most probably the same as the old town of Dharani-kotta, on the Kistna river, adjoining the modern city of Amaravati. Tse-kia., therefore, represents Ṭāki, which would appear to have been the name of the capital as well as of the kingdom of the Panjab in the seventh century, just as Lahor has since been used to describe both the kingdom and the
1 See Maps Nos. V. and VI.
[p.149]: capital of Ranjit Singh. The position of the capital will be discussed hereafter. It will be sufficient at present to note that it was within a few miles of the more ancient capital of She-kie-lo, which was long ago identified by Professor Lassen with the Sakala of the Mahabharata, and with the Sangala of Arrian.
Now the people of Sakala are called Madras, Arattas, Jarttikas, and Bahikas1 in the Mahabharata ; and in the Lexicon of Hemachandra the Bahikas are said to be the same as the Takkas (टक्का).2 Again, in the 'Raja Tarangini,' the district of Takkadesa is mentioned as a part of the kingdom of Gurjjara (or Gujrat, near the Chenab), which Raja Alakhana was obliged to cede to Kashmir between A.D. 883 and 901.3 From these statements it is clear that Sakala was the old capital of the powerful tribe of Takkas, whose country was named after themselves Takka-desa.4 The name of the new capital is not actually stated by Hwen Thsang, but I believe it to have been Taki, or Takkawar, which I would identify with the Tahora of the Pentingerian Tables by the mere softening of the guttural k to the aspirate h. In the latter authority Tahora is placed at 70 Roman miles, or 64⅓ English miles from Spatura, opposite Alexandria Bucefalos.
2 Lassen, ' Pentapot Indica,' p. 21. Bāhīkāshṭakkanāmāno.
3 'Raja Tarangini,' v. 150, Troyer; v. 155, Calcutta edit.
" The Mihran of es-Sind comes from the well-known sources of the high land of es-Sind, from the country belonging to Kinnauj in the kingdom of Budah, and of Kashmir, el Kandahar, and et-Takin. The tributaries which rise in these countries run to el Multan, and from thence the united " river receives the name of Mihran."
In this passage Takin must certainly be intended for the hills of the Panjab. The Kabul river and the Indus both flow through Gandhara or el Kandahar ; the Jhelam comes from Kashmir ; and the Bias and Satlej flow through Jalandhar and Kahlur, which in the time of Hwen Thsang were subject to Kanoj. The only other tributaries of the Indus are the Chenab and the Ravi, which must therefore have flowed through the kingdom of Takin. The mention of Gandhara and Kanoj shows that Masudi does not refer to the actual sources of the rivers, but to the points in the lower ranges of hills, where they enter the plains. Takin therefore, in the time of Masudi, represented the lower hills and plains of the Panjab to the north of Multan, which was then in the possession of the Brahman kings of Kabul.
The name is read Ṫākin, <arabic>, by Sir Henry Elliot, and Tafan, <arabic>, by Gildemeister,2 in his extracts from Masudi. The first reading is supported by the strong authority of Abu Rihan and Rashid-ud-
1 Sir H. M. Elliot's ' Muhammadan Historians of India,' p. 56; and Prof. Dowson's edition, i. 21, where the name is read as Tafan. But Sprenger, in his translation of 'Masudi,' p. 193, gives Tafi, with Takan and Tafan as variants, and at p. 390, Takin.
2 ' De Rebus Indicis,' p. 161.
[p.151]: din, who agree in stating that the great snowy mountain of Kelarjik (or Larjk), which resembled Demavend by its cupola form, could be seen from the boundaries of Takishar and Lohawar.1 Elliot, in one passage, corrects Takishar to Kashmir ; but this alteration is quite inadmissible, as the mountain is specially noted to have been only 2 farsangs, or about 8 miles, distant from Kashmir. One might as well say that St. Paul's Cathedral is visible from Ludgate Hill and Windsor. The mountain here referred to is the great Dayamur, or Nanga Parbat, to the west of Kashmir, which is 26,629 feet in height ; and which I have myself seen repeatedly from Ramnagar, on the Chenab, a distance of 200 miles. In a second passage of the same author, Sir Henry calls the mountain Kalarchal,2 and the two places from which it can be seen be names Takas and Lohawar. This Takas, or Takishar, I take to be the same place as the Tsekia, or Taki of Hwen Thsang, and the Takin of Masudi.
The earliest Muhammadan author who mentions Taki is the merchant Suliman, who visited the east before A.D. 851, when his account was written. He describes Tafak, <arabic>, as not of very great extent, and its king as weak, and subject to the neighbouring princes ; but he adds that he possessed "the finest white women in all the Indies.3 As Tafak and Takin are almost the same in unpointed Persian characters, I have
1 Reinaud, ' Fragments Arabes,' p. 118. In Sir H. M. Elliot, p. 41, and in Dowson's edition of Elliot, i. 65, Takishar is altered to Kashmir.
2 Sir H. M. Elliot, p. 30 ; and Dowson's edition, i. 46. If this is the same as Ibn Batuta's Karachal, or " Black Mountain," the identification with Nanga Parbat, or the " Bare Mountain " is nearly certain, as " bareness " means " blackness," from want of snow.
3 Sir Henry Elliot, p. 49 ; and Dowson's edition, i. 4.
[p.152]: no hesitation in identifying Tafak with the Panjab, where the women, and especially those of the lower hills, are the " fairest," as well as the " finest," in India.
Ibn Khurdadba, who died in A.D. 912, mentions the king of Taffa1 as next in eminence to the Balhara. Lastly, Kazwini describes Taifand, which was taken by Mahmud of Ghazni in A.D. 1023, as a strong Indian fort, on the top of an inaccessible mountain.2 This account agrees with the actual hill of Sangala, which is almost inaccessible on three sides, and on the fourth is protected by a sheet of water.
All these slightly different names of Takin, Tafan, Tafak, Taffa, Takas, and Takishar, I take to be only various readings of the one original form of Taki, or Takin, which, when written without the diacritical points, may be read in several different ways. M. Reinaud gives another spelling as Taban, which, without the points, may be read in as many different ways as the other form of Tafan. I conclude, therefore, that the true form of the name of the country was Taki (टाकी), or Taka (टाका), as recorded by Hwen Thsang. The name of the capital was probably either Takin or Takkawar, of which the former agrees exactly with Kazwini's Taifand, and the latter with the Tahora of the Pentingerian Tables. I consider it almost certain that the name must have been derived from the tribe of Taks or Takkas, who were once the undisputed lords of the Panjab, and who still exist as a numerous agricultural race in the lower hills between the Jhelam and the Ravi.
2 Gildemeister, ' De Rebus Indicis,' p. 208.
[p.153]: Ṭākari characters: The former importance of this race is perhaps best shown by the fact that the old Nāgari characters, which are still in use throughout the whole country from Bamiyan to the banks of the Jumna, are named Ṭākari, most probably because this particular form was brought into use by the Taks or Takkas. I have found these characters in common use under the same name amongst the grain dealers to the west of the Indus, and to the east of the Satlej, as well as amongst the Brahmans of Kashmir and Kangra. It is used in the inscriptions, as well as upon the coins of Kashmir and Kangra ; it is seen on the Sati monuments of Mandi, and in the inscriptions of Pinjor ; and lastly, the only copy of the ' Raja Tarangini ' of Kashmir was preserved in the Takari characters. I have obtained copies of this alphabet from twenty-six different places between Peshawar and Simla. In several of these places the Takari is also called Munde and Lunde, but the meaning of these terms is unknown. The chief peculiarity of this alphabet is, that the vowels are never attached to the consonants, but are always written separately, with, of course, the single exception of the inherent short a. It is remarkable also that in this alphabet the initial letters of the cardinal numbers have almost exactly the same forms as the nine unit figures in present use.
Kingdom of Taki: In the seventh century the kingdom of Taki was divided into three provinces, namely, Taki in the north and west, Shorkot in the east, and Multan in the south. The province of Taki comprised the plains of the Panjab, lying between the Indus and the Bias, to the north of the Multan district, or the whole of the Chaj Doab, together with the upper portions of the
[p.154]:three Doabs of Sindh-Sagar, Richna, and Bari. The province of Shorhot comprised the middle portions of these Doabs, and the province of Multan their lower portions. It is probable, also, that the possessions of Multan may have extended some distance to the west of the Indus as well as to the east of the Satlej, as was the case in the time of Akbar.
1. Taki or Northern Punjab
The province of Taki contained several of the most celebrated places of ancient India ; some renowned in the wars of Alexander, some famous in Buddhist history, and others known only in the widely-spread traditions of the people. The following is a list of the most important of the ancient places, arranged according to their relative geographical positions from west to east. The names of the Doabs were invented by Akbar by combining the names of the including rivers. Thus, Chaj is an abbreviation of Chenab and Jhelam ; Richna of Ravi and Chenab ; and Bari of Bias and Ravi.
|Sindh-Sagar Doab||1. Jobnathnagar, or Bhira.|
|2. Bukephala, or Dilawar.|
|Chaj Doab||3. Nikaea or Mong|
|Richna Doab||5. Sakala, or Sangala.|
|6. Taki, or Asarur.|
|7. Narsingha, or Ransi.|
|8. Ammakatis, or Ambakapi.|
|Bari Doab||9. Lohawar, or Lahor.|
|10. Kusawar, or Kasur.|
|11. Chinapati, or Patti.|
The modern town of Bhira, or Bheda, is situated on the left, or eastern bank, of the Jhelam ; but on the opposite bank of the river, near Ahmedabad, there is a very extensive mound of ruins, called Old Bhira, or Jobnathnagar, the city of Raja Jobnath, or Chobnath. At this point the two great routes of the salt caravans diverge to Lahor and Multan ; and here, accordingly, was the capital of the country in ancient times ; and here also, as I believe, was the capital of Sophites, or Sopeithes, the contemporary of Alexander the Great. According to Arrian, the capital of Sopeithes was fixed by Alexander as the point where the camps of Kraterus and Hephsestion were to be pitched on opposite banks of the river, there to await the arrival of the fleet of boats under his own command, and of the main body of the army under Philip.1 As Alexander reached the appointed place on the third day, we know that the capital of Sophites was on the Hydaspes, at three days' sail from Nikaea for laden boats. Now Bhira is just three days' boat distance from Mong, which, as I will presently show, was almost certainly the position of Nikaea, where Alexander defeated Porus. Bhira also, until it was supplanted by Pind Dadan Khan, has always been the principal city in this part of the country. At Bhira 2 the Chinese pilgrim, Fa-Hian, crossed the Jhelam in A.D. 400 ; and against Bhira, eleven centuries later, the enterprising Baber conducted his first Indian expedition.
The classical notices of the country over which
- " Some writers place Kathaea and the country of Sopeithes, one of the monarchs, in the tract between the rivers (Hydaspes and Akesines) ; some on the other side of the Akesines and of the Hyarotes, on the confines of the territory of the other Porus, — the nephew of Porus, who was taken prisoner by Alexander, and call the country subject to him Gandaris".
This name may, I believe, be identified with the present district of Gundalbar, or Gundar-bar. Bar is a term applied only to the central portion of each Doab, comprising the high lands beyond the reach of irrigation from the two including rivers. Thus Sandal, or Sandar-bar, is the name of the central tract of the Doab between the Jhelam and the Chenab. The upper portion of the Gundal Bar Doab, which now forms the district of Gujrat, belonged to the famous Porus, the antagonist of Alexander, and the upper part of the Sandar-Bar Doab belonged to his nephew, the other Porus, who is said to have sought refuge among the Gandaridae. The commentators have altered this name to Gangaridae, or inhabitants of the Ganges ; but it seems to me that the text of Diodorus2 is most probably correct, and that the name of Gandaridae must refer to the people of the neighbouring district of Gandaris, who were the subjects of Sophites. The rule of the Indian prince was not, however, confined to the Doab between the Hydaspes and Akesines; for Strabo: relates that "in the territory of
1 Geogr., XV. 1, 30. 2 Hist., xix. 47.
3 Geogr., XV. 1-30. This notice was most probably derived from Kleitarchoa, one of the companions of Alexander, as Strabo quotes him in another place (v. 2-6) as having mentioned the salt mines of India, <greek>.
[p.157]: Sopeithes there is a mountain composed of fossil salt sufficient for the whole of India." As this notice can only refer to the well-known mines of rock salt in the Salt Range, the whole of the upper portion of the Sindh Sagar Doab must have been included in the territories of Sopeithes. His sway, therefore, would have extended from the Indus on the west to the Akesines on the east, thus comprising the whole of the present districts of Pind Dadan and Shahpur. This assignment of the valuable salt mines to Sopeithes, or Sophites, may also be deduced from a passage in Pliny by the simple transposition of two letters in the name of a country, which has hitherto puzzled all the commentators. Pliny says, " when Alexander the Great was on his Indian expedition, he was presented by the king of Albania with a dog of unusual size," which successfully attacked both a lion and an elephant in his presence.1 The same story is repeated by his copyist, Solinus,2 without any change in the name of the country. Now, we know from the united testimony of Strabo, Diodorus, and Curtius, that the Indian king who presented Alexander with these fighting dogs was Sophites, and he, therefore, must have been the king of Albania. For this name I propose to read Labania, by the simple transposition of the first two letters. AABAN would, therefore, become AABAN, which at once suggests the Sanskrit word lavana, or ' salt,' as the original of this hitherto puzzling name. The mountain itself is named Orumenus by Pliny,3 who notes that the kings of the country
1 Hist. Nat., viii. 61.
2 Ibid., xxxi. 39. " Sunt et montes nativi salis, ut in Indis Oro- menus.
[p.158]: derived a greater revenue from the rock salt than from either gold or pearls. This name is probably intended for the Sanskrit Raumaka, which, according to the Pandits, is the name of the salt brought from the hill country of Ruma. H. H. Wilson, however, identifies Ruma with Sambhar ;1 and as rauma means " salt," it is probable that the term may have been applied to the Sambhar lake in Rajasthan, as well as to the Salt Range of hills in the Panjab.2
The historians of Alexander have preserved several curious particulars regarding Sophites and the country and people over which he ruled. Of the king himself, Curtius3 records that he was pre-eminent amongst the barbarians for beauty ; and Diodorus4 adds, that he was six feet in height. I possess a coin of fine Greek workmanship, bearing a helmeted head on one side, and on the reverse a cock standing, with the legend ΣΩΦΥΤΟΥ, which, there seems good reason to believe, must have belonged to this Indian prince. The face is remarkable for its very striking and peculiar features. The subjects of Sophites also were distinguished by personal beauty, which, according to Diodorus, they endeavoured to preserve, by destroying all their children who were not well formed. Strabo relates the same thing of the Katheai, but, as he adds, that they elected the handsomest person for their king,5 his account must be referred to the subjects of Sophites, as the Katheai of Sangala had no king. There is, however, so much confusion between all the authorities in their accounts of the Katheai and
2 See Maps Nos. V. and VI. 3 Vita Alex., ix. 1.
4 Hist., xvii. 49. 5 Geogr., xv. 1, .30.
[p.159]: of the subjects of Sophites, that it seems highly probable that they were one and the same people. They were certainly neighbours ; and as both of them would appear to have had the same peculiar customs, and to have been equally remarkable for personal beauty, I conclude that they must have been only different tribes of the same race of people,
The scene of Alexander's battle with Porus has long engaged the attention, and exercised the ingenuity, of the learned. The judicious Elphinstone1 placed it opposite to Jalalpur ; but Burnes2 concluded that it must have been near Jhelam, because that place is on the great road from Tartary, which appears to have been followed by Alexander. In 1836 the subject was discussed by General Court,3 whose early military training, and unequalled opportunities for observation during a long residence in the Panjab, gave him the best possible means of forming a sound opinion. General Court fixed the site of Alexander's camp at Jhelam, his passage of the river at Khilipatam, 3 kos, or 6 miles, above Jhelam, the scene of his battle with Porus at Pattikoti on the Jaba Nadi, 8 miles to the east of Jhelam, and the position of Nikaea at Vessa, or Bhesa, which is 3 miles to the south-east of Pathi or Patti-koti. The late Lord Hardinge took great interest in the subject, and twice conversed with me about it in 1846 and 1847. His opinion agreed with mine that the camp of Alexander was most
1 Elphinstone's ' Kabul,' i. 109.
2 ' Travels in Panjab, Bokhara, etc.,' ii. 49.
3 'Journal of the Asiatic Society,' Bengal, 1836, pp. 472, 473.
[p.160]: probably near Jalalpur. In the following year, General Abbott1 published an elaborate disquisition on the battle-field of Alexander and Porus, in which he placed the camp of the former at Jhelam, and of the latter on the opposite bank near Norangabad. The passage of the river he fixed at Bhuna, about 10 miles above Jhelam, and the field of battle near Pakrāl, about 3 miles to the north of Sukchenpur. In this state the question remained until the end of 1863, when my tour through the Panjab gave me an opportunity of examining at leisure the banks of the Hydaspes from Jalalpur to Jhelam.
Before discussing Alexander's movements, I think it best to describe the different places on the line of the river, between Jhelam and Jalalpur, with the approaches to them from the westward. When we have thus ascertained the site that will best agree with the recorded descriptions of Bukephala, we shall then be in a better position for deciding the rival claims of Jhelam and Jalalpur as the site of Alexander's camp. The distances that I shall make use of in this discussion are all taken from actual measurements.1
The town of Jhelam is situated on the west bank of the river, 30 miles to the north-east of Jalalpur, and exactly 100 miles to the north-north-west of Lahor. The remains of the old town consist of a large ruined mound, to the west of the present city, about 1300 feet square and 30 feet high, which is surrounded by fields covered with broken bricks and pottery. The square mound I take to be the ruins of the citadel, which is said to have been called Puta. Numbers of
1 ' Journal of the Asiatic Society,' Bengal, 1848, part ii. p. 619.
2 See No. VII. Map of Alexander's Passage of the Hydaspes,'
[p.161]: old coins are still discovered in the mound after rain ; but those which I was able to collect were limited to the mintages of the later Indo-Scythians, the Kabul-Brahmans, and the princes of Kashmir. As similar and even earlier coins are described by Court and Abbott to have been found in great numbers in previous years, it is certain that the city must have been in existence as early as the first century before Christ. But the advantages of its situation, on one of the two principal lines of road across the North Panjab, are so great that it must, I think, have been occupied at a very early date. This opinion is confirmed by the numbers of large bricks that have been dug out of the old mound.
The ruined city near Dārāpur, which has been described by Burnes1 and Court,2 is situated on the west bank of the river, 20 miles below Jhelam, and 10 miles above Jalalpur. In their time, the old mound was unoccupied, but about 1832 A.D. the people of Dilawar abandoned their village on a hill to the west, and settled on the site of the ruined city. Before that time, the place was usually called Find, or " the mound," although its true name is said to have been Udamnagar, or Udinagar. The same name is also given by Burnes, but Court, who twice alludes to these ruins, mentions no name, unless he includes them under that of Gagirakhi, the ruins of which he describes as extending along the banks "of the Hydaspes from near Jalalpur to Darapur." According to this account, the ruins would not be less than 6 or 7 miles in length. I think it probable that there has
1 ' Travels in Panjab, Bokhara, etc.,' ii. 51. 2 Journ. Asiat. Soc. Bengal, 1836, 472, 473.
[p.162]: been some confusion between two different places, which have here been joined together as one continuous extent of ruins. Girjhāk, which I take to be the original of Court's Gagirakhi, is an old ruined fort on the top of the hill to the north of Jalalpur, to which the people assign a fabulous extent ; but it is at least 8 miles from Dardpur, and is, besides, separated from it by the deep Kandar ravine, and by the precipitous range of hills at whose west foot Dilawar is situated. Burnes also describes the old city as extending "for three or four miles." But this is certainly an exaggeration, as I was unable to trace the ruins for more than one mile in length by half a mile in breadth. The ruins consist of two large mounds just half a mile apart, with two smaller mounds about midway between them. The south mound on which Dilawar is situated, is about 500 feet square at top, and 1100 or 1200 feet at base, with a height of 50 or 60 feet. The north mound, on which old Darapur stands, is 600 feet square, and from 20 to 30 feet in height. Between these mounds the fields are covered with broken bricks and pottery, and the whole place is said to be the ruins of a single city. The walls of the Dilawar houses are built of the large old bricks dug out of this mound, which are of two sizes, one of 11½ by 8¼ by 3 inches, and the other of only half this thickness. Old coins are found in great numbers in the Dilawar mound, from which the Jalalpur bazar is said to be supplied, just as Pind Dadan is supplied from the ruins of Jobnathnagar. The coins which I obtained belonged to the first Indo-Scythians, the Kabul-Brahmans, the kings of Kashmir, and the Karluki Hazara chiefs, Hasan and his son Muhammad. The site.
[p.163]: therefore, must have been occupied certainly as early as the second century before the Christian era. Its foundation is attributed to Raja Bharati, whose age is not known. I conclude, however, that the dominating position of Dilawar, which commands the passage of the Jhelam at the point where the lower road from the west leaves the hills, just below the mouth of the Bunhar river, must have led to its occupation at a very early period.
Position of Jalalpur
The town of Jalalpur is situated on the west bank of the Jhelam at the point where the Kandar ravine joins the old bed of the river. The stream is now 2 miles distant, but the intervening ground, though partially covered with small trees, is still very sandy. The town is said to have been named in honour of Akbar, in whose time it was most probably a very flourishing place. But since the desertion of the river, and more especially since the foundation of Pind Dadan, the place has been gradually decaying, until it now contains only 738 houses, with about 4000 inhabitants. From the appearance of the site, I estimated that the town might formerly have been about three or four times its present size. The houses are built on the last slope at the extreme east end of the salt range, which rises gradually to a height of 150 feet above the road. Its old Hindu name is said to have been Girjhak, and as this name is found in Abul Fazl's ' Ayin Akbari'1 as Kerchak (read Girjak) of Sindh Sagar, we have a proof that it was in use until the time of Akbar, when it was changed to Jalalpur. But the people still apply the name of Girjhak to the remains of walls on the top of the Mangal-De hill,
1 Gladwyn's Translation, ii. 263.
[p.164]: which rises 1100 feet above Jalalpur. According to tradition, Girjhak extended to the west-north-west as far as the old temple of Baghanwala, a distance of 11 miles. But this is only the usual exaggeration of ignorance that is told of all ancient sites. There is no doubt that the city did once extend to the westward for some considerable distance, as the ground on that side is thickly strewn with broken pottery for about half a mile.
Its antiquity is undoubted, as the coins which it yields reach back to the times of Alexander's successors. But I believe that it is much older, as its favourable position at the south-east end of the lower road would certainly have led to its occupation at a very early period. I think, therefore, that it may be identified with the Girivraja of the Ramayana. Tradition has preserved the name of only one king, named Kumkamarath, who is said to have been the sister's son of Moga, the founder of Mong. Mogal Beg1 writes the name Ghir-Jehak, and it is so written by some of the people of the place, as if it was derived from Giri-Zohak, or " Zohak's Hill." But the usual spelling, which accords with the pronunciation, is Jhak. From Jhelam to Jalapur the course of the river is from north-east to south-west, between two nearly parallel ranges of mountains, which are generally known as the Tila and Pabhi Hills. The Tila range, which is about thirty miles in length, occupies the west bank from the great east bend of the river below Mangala, to the bed of the Bunhar river, 12 miles to the north of Jalapur. Tila means simply a "peak or hill," and the full name is Gorakhnath-ka-Tila. The more ancient
1 Manuscript Map of the Panjab and Kabul Valley, by Wilford, from the surveys of Mirza Mogal Beg, in my possession.
[p.165]: name was Balnath-ka-Tila. Both of these are derived from the temple on the summit, which was formerly dedicated to the sun, as Balnath, but is now devoted to the worship of Gorakhnath, a form of Siva. The latter name, however, is very recent, as Mogal Beg, who surveyed the country between A.D. 1784 and 1794, calls the hill Jogion-di-Tibi, or tower of the Jogis, whose chief is called Bilnath. Abul Fazl1 also mentions the " Cell of Balnat," and the attendant Jogis, or devotees, from whom the hill is still sometimes called Jogi-tila. But the name of Balnath is most probably even older than the time of Alexander, as Plutarch2 relates that, when Porus was assembling his troops to oppose Alexander, the royal elephant rushed up a hill sacred to the Sun, and in human accents exclaimed,
- " great king, who art descended from Gegasios, forbear all opposition to Alexander, for Gegasios himself was also of the race of Jove."
The " Hill of the Sun " is only a literal translation of Balnath-ka-Tila, but Plutarch goes on to say that it was afterwards called the " Hill of the Elephant," which I take to be another proof of its identity with Balnath, for as this name is commonly pronounced Bilnat by the people, and is so written by Mogal Beg, the Macedonians, who had just come through Persia, would almost certainly have mistaken it for Fil-nath, or Pil-nath, the "Elephant." But wherever Alexander's camp may have been, whether at Jhelam or Jalalpur, this remarkable hill, which is the most commanding object within fifty miles of the Hydaspes,
1 'Ayin Akbari,' ii. 110.
[p.166]: must certainly have attracted the attention of the Macedonians. Its highest peak is 3242 feet above the sea, or about 2500 feet above the level of the river.
The Pabhi range of hills, on the east bank of the river, stretches from the neighbourhood of Bhimbar to Rasul, a length of 30 miles. This range is a very low one, as the highest point is not more than 1400 feet above the sea, and is less than 500 feet above the river; but the broken and difficult ground on both flanks of the hill presents a barrier quite as impassable as a much loftier range. Until the British occupation of the Panjab, the Pabhi hills were crossed by only one carriage-road through the Khori Pass, 5 miles to the north-east of Rasul, and by one foot-path through the Kharian Pass, 10 miles to the south-east of Jhelam. But though the main road has since been carried through the latter pass, it is still liable to interruption after heavy rain.
In approaching the Hydaspes from the westward, Alexander had the choice of two different lines, which are distinguished by Baber as the upper and lower roads. From the Indus to Hasan Abdal, or Shah-dheri, the two lines were the same. From the latter place, the upper road proceeded by the Margala Pass through Rawal Pindi and Manikyala to Dhamak and Bakrala, from which place it descended by the bed of the Kahan river, through a gap in the Tila range, to Rohtas, and from thence over an open plain to Jhelam. From Bakrala there was also a foot-path to Jhelam, which crossed the Tila range about 6 miles to the north-east of Rohtas, but this pass was always a dangerous one for horses and camels, and was difficult even for foot passengers. The length of this
From Taxila, or Shah-dheri, the lower road proceeds via the Margala Pass to Jangi, from whence it strikes off via Chaontra to Dudhial. From this point the road branches into two lines, that to the south proceeding by Chakowal and the salt mines to Pind Dadan and Ahmadabad, and that to the east proceeding via Asanot and the Bunhar river to Dilawar, opposite Rasul, or via Asanot and Vang to Jalalpur. From Shah-dheri to Dudhial the distance is 55 miles, from thence to Asanot 33 miles, and thence -to Dilawar, or Jalalpur, each 21 miles, the whole distance by this route being 118 miles. But this distance would be shortened to 114 miles by the traveller proceeding direct from the foot of the Salt Range to Jalalpur. There is also a third line, which branches off from the upper road at Mandra, 6 miles to the south of the Manikyala tope, and proceeds via Chakowal and Pind Dadan to Jalalpur. By this route the whole distance from Shah-dheri to Jalalpur is 116¾ miles, or only 112¾ by leaving the line at the foot of the Salt Range and proceeding direct to Jalalpur. The respective distances by these three different routes are 109, 114, and 112¾ miles, the mean distance being 112¼ miles.
Now, the distance from Taxila to the Hydaspes is given by Pliny,1 from the measurement by Alexander's surveyors, Diognetes and Beiton, at 120 Roman miles, which are equal to 110⅓ English miles, at the value of 0-9193 each, as fixed in Smith's
1 Hist. Nat.i vi. 21, "Ad Hydaspen fluvium clarum, cxx. mill."
[p.168]: 'Dictionary of Antiquities.' As all the copies of Pliny give the same number, we must accept it as the actual measurement of the route that was followed by Alexander from Taxila to his camp on the Hydaspes. In comparing this distance with those already given from Shah-dheri to Jhelam and Jalalpur, we must unhesitatingly reject Jhelam, which is no less than 16 miles short of the recorded distance, while Jalalpur differs from it by less than 2 miles. But there is another objection which is equally fatal to the claims of Jhelam. According to Strabo,1 "the direction of Alexander's march, as far as the Hydaspes, was, for the most part, towards the south ; after that, to the Hypanis, it was more towards the east." Now, if a line drawn on the map from Ohind on the Indus, through Taxila to Jhelam, be continued onwards, it will pass through Gujarat and Sodhra to Jalandhar and Sarhind. As this is the most northerly road to the Ganges that Alexander could possibly have taken, his route by Jhelam would have been in one continuous straight line, which is in direct opposition to the explicit statement of Strabo. But if we adopt Jalalpur this difficulty will be obviated, as the change in the direction would have been as much as 25° more easterly.2 There is also a third objection to Jhelam, which, though not entitled to the same weight as either of the preceding, is still valuable as an additional testimony on the same side. According to Arrian, the fleet, on descending the Hydaspes from Nikaea, reached the capital of Sopeithes on the third day. Now, I have already shown that the residence of Sopeithes must have been at Jobnathnagar,
1 Geogr., XV. 1, 32. 2 See Map No. V.
or Ahmedabad, which is just three days' distance for a laden boat from Jalalpur, but is six days from Jhelam. As the evidence in each of these three separate tests is as directly in favour of Jalalpur as it is strongly opposed to Jhelam, I think that we are fully justified in accepting the latter as the most probable site of Alexander's camp.
We have now to examine how the river and the country about Jalalpur will agree with the recorded accounts of Alexander's operations in his passage of the Hydaspes and subsequent battle with Porus. According to Arrian1 " there was a high wooded promontory on the bank of the river, 150 stadia, or just 17¼ miles above the camp, and immediately opposite to it there was a thickly-wooded island." Curtius2 also mentions the wooded island as "well fitted for masking his operations." "There was also," he adds,"not far from the spot where he was encamped, a very deep ravine [fossa praealia), which not only screened the infantry but the cavalry too." We learn from Arrian3 that this ravine was not near the river because " Alexander marched at some distance from the bank, lest the enemy should discern that he was hastening towards the promontory and island." Now, there is a ravine to the north of Jalalpur which exactly suits the descriptions of both historians. This ravine is the bed of the Kandar Nala, which has a course of 6 miles from its source down to Jalalpur, where it is lost in a waste of sand. Up this ravine
1 ' Anabasis,' v.11. <greek>.
2 Vita Alez., viii. 13, " tegendis insidiis apta."
3 'Anabasis,' v. 12, <greek>.
[p.170]: there has always been a passable but difficult road towards Jhelam. From the head of the Kandar, which is 1080 feet above the sea, and 345 above the river, this road proceeds for 3 miles in a northerly direction down another ravine called the Kasi, which then turns suddenly to the east for 6½ miles, and then again 1½ mile to the south, where it joins the Jhelam immediately below Dilawar, the whole distance from Jalalpur being exactly 17 miles. I marched along this ravine road myself, for the purpose of testing the possibility of Alexander's march ; and I satisfied my-self that there was no difficulty in it except the fatigue of making many little ascents and descents in the first half, and of wading through much heavy sand in the latter half. The ravine lies " at some distance from the bank " as described by Arrian, as the bend in the Kasi is 7 miles from the Jhelam. It is also " a very deep ravine," as described by Curtius, as the hills on each hand rise from 100 to 250 and 300 feet in height. Therefore, in the three leading particulars which are recorded of it, this ravine agrees most precisely with the accounts of the ancient historians,1 Amongst the minor particulars, there is one which seems to me to be applicable only to that part of the river immediately above Jalalpur. Arrian2 records that Alexander placed running sentries along the bank of the river, at such distances that they could see each other, and communicate his orders. Now, I believe that this operation could not be carried out in the face of an observant enemy along any part of the river
1 See Map No. VII.
2 ' Anabasis,' v. ii. <greek>
[p.171]: bank, excepting only that one part which lies between Jalalpur and Dilawar. In all other parts, the west bank is open and exposed, but in this part alone the wooded and rocky hills slope down to the river, and offer sufficient cover for the concealment of single sentries. As the distance along the river bank is less than 10 miles, and was probably not more than 7 miles from the east end of the camp, it is easy to understand why Alexander placed them along this line instead of leaving them on the much longer route, which he was to march himself. Another minor particular is the presence of a rock in the channel by the river, on which, according to Curtius, one of the boats was dashed by the stream. Now, rocks are still to be found in the river only at Kotera, Meriala, Malikpur, and Shah Kubir, all of which places are between Dilawar and Jalalpur. The village of Kotera is situated at the end of a long wooded spur, which juts out upon the river just one mile below Dilawar. This wooded jutting spur, with its adjacent rock, I would identify with the akρa, or promontory of Arrian, and the petra of Curtius.1 Beyond the rock there was a large wooded island which screened the foot of the promontory from the observation of the opposite bank. There are many islands in this part of the Jhelam, but when a single year is sufficient to destroy any one of these rapidly formed sandbanks, we can not, after the lapse of more than 2000 years, reasonably expect to find the island of Alexander. But in 1849, opposite Kotera, there was such an island,
[p.172]: 2½ miles in length and half a mile in breadth, which still exists as a large sandbank. As the passage was made in the height of the rainy season, the island, or large sandbank, would naturally have been covered with tamarisk bushes, which would have been sufficiently high to screen the movements of infantry and dismounted cavalry.
The position of the two camps I believe to have been as follows :1 — Alexander, with about 50,000 men, including 5000 Indian auxiliaries under Mophis of Taxila, had his head-quarters at Jalalpur, and his camp probably extended for about 6 miles along the bank of the river, from Shah Kabir, 2 miles to the north-east of Jalalpur, down to Syadpur, 4 miles to the west-south-west. The head- quarters of Porus must have been about Muhabatpur, 4 miles to the west-south-west of Mong, and 3 miles to the south- east of Jalalpur. His army of nearly 50,000 men, including elephant-riders, archers, and charioteers, must have occupied about the same extent as the Macedonian army, and would, therefore, have extended about 2 miles above, and 4 miles below Muhabatpur. In these positions, the left flank of Alexander's camp would have been only 6 miles from the wooded promontory of Kotera, where he intended to steal his passage across the river, and the right flank of the Indian camp would have been 2 miles from Mong, and 6 miles from the point opposite Kotera.
As my present object is to identify the scene of Alexander's battle with Porus, and not to describe the fluctuations of the conflict, it will be sufficient to quote the concise account of the operation which is given by Plutarch from Alexander's own letters: — "He took
- Sec Map No. VII.
[p.173]: advantage of a dark and stormy night, with part of his infantry and a select body of cavalry, to gain a little island in the river, at some distance from the Indians ; when he was there, he and his troops were attacked with a most violent wind and rain, accompanied with dreadful thunder and lightning." But in spite of the storm and rain, they pushed on, and wading through the water breast-high reached the opposite bank of the river in safety. " When they were landed," says Plutarch,1 who is still quoting Alexander's letters, " he advanced with the horse 20 stadia before the foot, concluding, that if the enemy attacked him with their cavalry he should be greatly their superior, and that if they made a movement with their infantry his own would come up in time enough to receive them." From Arrian,2 we learn that, as soon as the army had begun fording the channel, between the island and the main land, they were seen by the Indian scouts, who at once dashed off to inform Porus. When the ford was passed with some difficulty, Alexander halted to form his little army of 6000 infantry and about 10,000 cavalry. He then " marched swiftly forward with 5000 horse, leaving the infantry to follow him leisurely and in order." While this was going on, Porus had detached his son with two or three thousand horse and one hundred and twenty chariots to oppose Alexander. The two forces met at 20 stadia, or 2¼ miles, from the place of crossing, or about two miles to
1 ' Life of Alexander.' Sir W. Napier has paid a just tribute to the skill of both generals. Speaking of Alexander's passage of the Granicus, he says that it cannot " be compared for soldierly skill with his after passage of the Hydaspes, and defeat of Porus. Before that great man he could not play the same daring game." (' London and Westminster Review,' 1838, p. 377.)
2 ' Anabasis,' v. 18.
[p.174]: the north-east of Mong. Here the chariots proved useless on the wet and slippery clay, and were nearly all captured. The conflict, however, must have been a sharp one, as Alexander's favourite charger, Bukephalus, was mortally wounded by the young prince, who was himself slain, together with 400 of his men. When Porus heard of the death of his son, he marched at once against Alexander with the greater part of his army ; but when he came to a plain, where the ground was not difficult and slippery, but firm and sandy, and fitted for the evolutions of his chariots, he halted and arrayed his troops ready for battle. His 200 elephants were drawn up in front of the infantry about one pleihron, or 100 feet apart, and the chariots and cavalry were placed on the flanks. By this arrangement, the front of the army facing north-east must have occupied an extent of about 4 miles, from the bank of the river to near Lakhnawali, the centre of the line being, as nearly as possible, on the site of the present town of Mong. Around this place the soil is " firm and sound ; " but towards the north-east, where Alexander encountered the young Indian prince, the surface is covered with a hard red clay, which becomes both heavy and slippery after rain.1
When Alexander saw the Indian army drawn up in battle array, he halted to wait for his infantry, and to reconnoitre the enemy's position. As he was much superior to Porus in cavalry, he resolved not to attack the centre, where the formidable line of
1 I speak from actual observation of the field of Chilianwala for some days after the battle, when the country had been deluged with rain. Both battles were fought on the same ground, between the town of Mong and the southern end of the Pabhi hills.
[p.175]: elephants were supported by masses of infantry, but to fall upon both flanks and throw the Indians into disorder. The right wing, led by Alexander himself, drove back the enemy's horse upon the line of elephants, which then advanced and kept the Macedonians in check for some time. " Wherever Porus saw cavalry advancing, he opposed elephants, but these slow and unwieldy animals could not keep pace with the rapid evolutions of the horse."1 At length the elephants, wounded and frightened, rushed madly about, trampling down friends as well as foes. Then the small body of Indian horse being surrounded, was overpowered by the Macedonians, and nearly all slain ; and the large mass of Indian infantry, which still held out, being vigorously attacked on all sides by the victorious horse, broke their ranks and fled. Then, says Arrian2 " Kraterus, and the captains who were with him on the other side of the river, no sooner perceived the victory to incline to the Macedonians, than they passed over, and made a dreadful slaughter of the Indians in pursuit."
From the last statement which I have quoted, it is clear that the battle-field was within sight of Alexander's camp. Now, this is especially true of the plain about Mong, which is within easy ken of the east of Alexander's camp at Shah-Kabir, the nearest point being only 2 miles distant. With this last strong evidence in favour of Jalalpur as the site of Alexander's camp, I close my discussion of this interesting question. But as some readers, like Mr. Grote,3 the historian of Greece, may still think that
1 Curtius, Vita Alex., viii. 14, 27.
2 ' Anabasis,' v. 18.
3 ' History of Greece,' xii. 308, note.
[p.176]: General Abbott has shown " highly plausible reasons " in support of his opinion that Alexander's camp was at Jhelam, I may here point out that the village of Pabrāl, which he has selected as the battle-field, is not less than 14 miles from Jhelam, and therefore quite beyond the ken of Alexander's camp. I may quote also Abbott's own admission that the bed of the Sukhetr river, a level plain of sand one mile in width, " is a torrent after heavy rain, and is so full of quick-sands as to be unsuited to military operations." Now, this very Sukhetr river actually lies between Pabral and the site of the Indian camp opposite Jhelam, and as we know that a heavy storm of rain had fallen during the preceding night, the Sukhetr would have been an impassable torrent at the time of the battle. And so also would have been the Jada river, which joins the Jhelam just below the Sukhetr. With these two intervening rivers, which, whether wet or dry, would have been obstacles equally great to the march of the Indian army, and more specially to the passage of the war-chariots, I am quite satisfied that the battle-field could not have been to the north of the Sukhetr river. The position of Bukephala still remains to be discussed.
Position of Bukephala
According to Strabo,1 the city of Bukephala was built on the west bank of the river, where Alexander had crossed it ; but Plutarch2 says that it was near the Hydaspes, in the place where Bukephalus was buried. Arrian,3 however, states that it was built on the site of his camp, and was named Bukephala in memory of his horse. Diodorus, Curtius, and Justin leave the exact position undecided ; but they all agree that it was on the opposite bank of the
1 Geogr., XV. 1, 29.
2 ' Life of Alexander.'
3 ' Anabasis,' v. 19.
[p.177]: river to Nikaea, which was certainly built on the field of battle. With these conflicting statements alone to guide us, it is diffcult to arrive at any positive conclusion. According as we follow Strabo or Arrian, we must place Bukephala at Dilawar, or at Jalalpur. Both places are equidistant from the battle-field of Mong, which I take without much hesitation to be the site of Nikaea. If the two cities were built on the same plan, which is not improbable, then Dilawar would have the preferable claim to represent Bukephala, as its ruined mound is of the same size and height as that of Mong. I have already noticed in another place the possibility that Bugiad, or Bugial, the name of the district in which Dilawar is situated, may be only an abbreviation of Bukephalia by the easy elision of the ph. But this is only a guess, and I have nothing else to offer on the subject, save the fact that the ancient name of Jalalpur was certainly Girjak, while that of Dilawar is quite uncertain, as Udinagar is applied to at least three different places. The claims of Dilawar and Jalalpur are perhaps equally balanced, excepting in the one important point of position, in which the latter has a most decided advantage ; and as this superiority would not have escaped the keen observation of the founder of Alexandria, I think that Jalalpur must be the site of the famous city of Bukephala.
The position of Mong has already been described, but I may repeat that it is 6 miles to the east of Jalalpur, and the same distance to the south of Dilawar. The name is usually pronounced Mong, or Mung
[p.178]: but it is written without the nasal, and is said to have been founded by Raja Moga, or Muga. He is also called Raja Sankhar, which I take to mean king of the Sakas, or Sakae. His brother Rama founded Rampur, or Ramnagar, the modern Rasul, which is 6 miles to the north-east of Mong, and exactly opposite Dilawar. His sister's son, named Kamkamarath, was Raja of Girjak or Jalalpur. The old ruined mound on which Mong is situated, is 600 feet long by 400 feet broad and 50 feet high, and is visible for many miles on all sides. It contains 975 houses built of large old bricks and 5000 inhabitants, who are chiefly Jats. The old wells are very numerous, their exact number, according to my informant, being 175.
I have already stated that I take Mong to be the site of Nikaea, the city which Alexander built on the scene of his battle with Porus. The evidence on this point is, I think, as complete as could be wished ; but I have still to explain how the name of Nikaea could have been changed to Mong. The tradition that the town was founded by Raja Moga is strongly corroborated by the fact that Maharaja Moga is mentioned in Mr. Roberts's Taxila inscription. Now, Moga is the same name as Moa, and the coins of Moa, or Mauas are still found in Mong. But the commonest Greek monogram on these coins forms the letters NIK, which I take to be the abbreviation of Nikaea, the place of mintage. If this inference be correct, as I believe it is, then Nikaea must have been the principal mint-city of the great king Moga, and therefore a place of considerable importance. As the town of Mong is traditionally attributed to Raja Moga as the founder, we may reasonably conclude that he must
[p.179]: have rebuilt or increased the place under the new name of Moga-grama, which, in the spoken dialects, would be shortened to Mogaon and Mong. Coins of all the Indo-Scythian princes are found at Mong in, considerable numbers, and I see no reason to doubt that the place is as old as the time of Alexander. The copper coins of the Nameless Indo-Scythian king especially are found in such numbers at Mong that they are now commonly known in the neighbourhood as Monga-sahis.
The city of Gujarat is situated 9 miles to the west of the Chenab river, on the high-road from Jhelam to Lahor. The city is said to have been first called Hairat, and the district Hairat-des.1 Its original foundation is ascribed to a Surajbansi Rajput named Bachan Pal, of whom nothing more is known ; and its restoration is attributed to Ali Khan, a Gujar, whose name is strangely like that of Alakhana, the Raja of Gurjjara, who was defeated by Sangkara Varmma between A.D. 883 and 901. Following up these traditions, Gujarat is said to have been destroyed in A.D. 1303, and to have been rebuilt by the Gujars in A.H. 996, or A.D. 1588, during the reign of Akbar.
The Sangala of Alexander has long ago been recognized in the Sakala of the Brahmans and the Sāgal of the Buddhists; but its position would still perhaps have remained undetermined, had it not fortunately been visited by the Chinese pilgrim Hwen Thsang in
[p.180]: A.D. 630. Both Arrian and Curtius place Sangala to the east of the Hydraotes, or Ravi ; but the itinerary of Hwen Thsang shows that it was to the west of the Ravi, and as nearly as possible in the position of the present Sangla-wala-Tiba, or " Sangala Hill." I first became acquainted with this place in 1839, when I obtained a copy of Mogal Beg's manuscript map, compiled by Wilford, who has three times described its position in the ' Asiatic Researches.'1 But I was not able to obtain any account of the place until 1854, when I heard from Colonel G. Hamilton, who had visited it, and from Captain Blagrave, who had surveyed it, that Sangala was a real hill with traces of buildings, and with a sheet of water on one side of it. During my tour through the Panjab, I was able to visit the hill myself, and I am now satisfied that it must be the Sangala of Alexander, although the position does not agree with that which his historians have assigned to it.
In the time of Hwen Thsang She-kie-lo, or Sakala, was in ruins, and the chief town of the district was Tse-kia, or Chekia, which may also be read as Dhaka or Taka. The pilgrim places this new town at 15 li, or 2½ miles, to the north-east of Sakala; but as all the country within that range is open and flat, it is certain that no town could ever have existed in the position indicated. In the same direction, however, but at 19 miles, or 115 li, I found the ruins of a large town, called Asarur, which accord almost exactly with the pilgrim's description of the new town of Tse-kia. It is necessary to fix the position of this place, because Hwen Thsang's measurements, both coming and going,
1 Vols. V. 282 ; vi. 520 ; ix. 53.
[p.181]: are referred to it and not to Sakala. From Kashmir the pilgrim proceeded by Punach to Rajapura, a small town in the lower hills, which is now called Rajaori. From thence he travelled to the south-east over a mountain, and across a river called Chen-ta-lo-po-kia, which is the Chandrabhaga, or modern Chenab, to She-ye-pu-lo, or Jayapura (probably Hafizabad), where he slept for the night, and on the next day he reached Tse-kia, the whole distance being 700 li, or 116 miles. As a south-east direction would have taken the pilgrim to the east of the ravi, we must look for some known point in his subsequent route as the best means of checking this erroneous bearing. This fixed point we find in She-lan-to-lo, the well-known Jalandhara, which the pilgrim places at 500, plus 50, plus 140 or 150 li, or altogether between 690 and 700 li to the east of Tse-kia. This place was, therefore, as nearly as possible, equidistant from Rajaori and Jalandhar. Now, Asarur is exactly 112 miles distant from each of these places in a direct line drawn on the map, and as it is undoubtedly a very old place of considerable size, I am satisfied that it must be the town of Tse-kia described by Hwen Thsang.
In AD. 630 the pilgrim found the walls of Sakala completely ruined, but their foundations still remained, showing a circuit of about 20 li, or 3⅓ miles. In the midst of the ruins there was still a small portion of the old city inhabited, which was only 6 or 7 li, or just one mile, in circuit. Inside the city there was a monastery of one hundred monks who studied the Hinayana, or exoteric doctrines of Buddhism, and beside it there was a stupa, 200 feet in height, where the four previous Buddhas had left their footprints.
[p.182]: At 5 or 6 li, or less than 1 mile, to the north-west, there was a second stupa, also about 200 feet high, which was built by King Asoka on the spot where the four previous Buddhas had explained the law.
Sanglawala Tiba is a small rocky hill forming two sides of a triangle, with the open side towards the south-east. The north side of the hill rises to a height of 215 feet, but the north-east side is only 160 feet. The interior area of the triangle slopes gradually down to the south-east till it ends abruptly in a steep bank 32 feet above the ground. This bank was once crowned with a brick wall, which I was able to trace only at the east end, where it joined the rock. The whole area is covered with brick ruins, amongst which I found two square foundations. The bricks are of a very large size, 15 by 9 by 3 inches. During the last fifteen years these bricks have been removed in great numbers. Nearly 4000 were carried to the large village of Marh, (1 miles to the north, and about the same number must have been taken to the top of the hill to form a tower for the survey operations. The base of the hill is from 1700 to 1800 feet on each side, or just 1 mile in circuit. On the cast and south sides the approach to the hill is covered by a large swamp, half a mile in length, and nearly a quarter of a mile in breadth, which dries up annually in the summer, but during the seasonal rains has a general depth of about 3 feet. In the time of Alexander this must have been a fine sheet of water, which has been gradually lessened in depth by the annual washings of silt from the hill above. On the north-eastern side of the hill there are the remains of two large buildings, from which I obtained old bricks of the enormous size
[p.183]: of 17½ by 11 by 3 inches. Close by there is an old well which was lately cleared out by some of the wandering tribes. On the north-western side, 1000 feet distant, there is a low ridge of rock called Munda-ka-pura, from 26 to 30 feet in height, and about 500 feet in length, which has formerly been covered with brick buildings. At 1¾ mile to the south, there is another ridge of three small hills, called Arna and little Sangala.
All these hills are formed of the same dark grey rock as that of Chanyot and of the Karana hills to the west of the Chenab, which contains much iron, but is not worked on account of the want of fuel. The production of iron is noticed by Hwen Thsang.
In comparing this account with the description of the Chinese pilgrim, I only find two places that can be identified. The first is the site of the modern town, which was about a mile in circuit, and was situated in the midst of the ruins. This I take to be the hill itself, which accords exactly with the description, and which would certainly have been occupied in preference to any part of the open plain below, on account of its security. The second is the stupa of Asoka, which was situated at rather less than 1 mile to the north-west of the monastery inside the town. This I would identify with the low ridge of rock on the north-west called Mundapapura, of which the highest point at the north-western end is 4000 feet, or more than three-quarters of a mile distant from the central point of the triangular area of the town. The plain on the north and west sides of the hill is strewn with broken pottery and fragments of brick for a considerable distance, showing that the town must once have extended in both of those directions. But the
[p.184]: whole circuit of these remains did not appear to be more than 1½ or 1¼ miles, or about one-half of Hwen Thsang's measurement.
The Brahmanical accounts of Sakala have been collected from the Mahabharata by Professor Lassen in his ' Pentapotamia Indica.'1 According to that poem, Sakala, the capital of the Madras, who are also called Jartikas and Bahikas, was situated on the Apaga rivulet to the west of the Iravati, or Ravi river. It was approached from the east side by pleasant paths through the Pilu forest,
which Professor Lassen translates " per amcenas sylvarum tramites ambulantes." But the Pilu, or Salvadora Persica, is the commonest wood in this part of the Panjab, and is specially abundant in the Rechna Doab. In these "pleasant paths" of the Pilu forest, the traveller was unfortunately liable to be despoiled of his clothes by robbers. This description by the author of the Mahabharata was fully verified by Hwen Thsang in A.D. 630, and again by myself in 1863. On leaving Sakala, the Chinese pilgrim travelled eastward into a forest of Po-lo-she trees, where his party encountered fifty brigands, who robbed them of their clothes.2 In November, 1863, I approached Sakala from the east through a continuous wood of Pilu trees, and pitched my tent at the foot of the hill. During the night the tent was three times approached by parties of robbers who were detected by the vigilance of my watch dog. M. Julien has properly rendered Hwen Thsang Po-lo-she by Palasa, the Butea frondosa,
1 Pentapot. Ind., pp. 73, 74. 2 ' Hiouen Thsang,' i. 97.
[p.185]: or Dhak tree ; but as the forest consisted of Pilu trees, both before and after the time of Hwen Thsang, I would suggest the propriety of correcting Pi-lo-she to Pilo ; I conjecture that the Chinese editor of the pilgrim's life, who was most probably ignorant of the Pilu, substituted the well-known Palasa, which is frequently mentioned by Hwen Thsang, under the belief that he was making an important and necessary correction.
The country is still well known as Madr-des, or the district of the Madras, which is said by some to extend from the Bias to the Jhelam, but by others only to the Chenab. Regarding the Apaga rivulet, I believe that it may be recognized in the Ayak Nadi, a small stream which has its rise in the Jammu hills to the north-east of Syalkot. After passing Syalkot the Ayak runs westerly near Sodhra, where in the rainy season it throws off its superfluous water in the Chenab. It then turns to the south-south-west past Banka and Nandanwa to Bhutala, and continues this same course till within a few miles of Asarur. There it divides into two branches, which, after passing to the east and west of Asarur, rejoin at 2½ miles to the south of Sangalawala Tiba. Its course is marked in the revenue survey maps for 15 miles to the south-west of Sangala, where it is called the Nananwa canal. An intelligent man of Asarur informed me that he had seen the bed of the Nananwa 20 kos to the south-west, and that he had always heard that it fell into the Ravi a long way off. This, then, must be Arrian's "small rivulet" near which Alexander pitched his camp, at 100 stadia, or 11½ miles, to the east of the Akesines, below its junction with the Hydaspes.1 At
1 ' Anabasis,' vi. 6.
[p.186]: that time, therefore, the water of the Ayak must have flowed for a long distance below Sangala, and most probably fell into the Ravi, as stated by my informant. Near Asarur and Sangala, the Ayak is now quite dry at all seasons ; but there must have been water in it at Dhakawala only 24 miles above Asarur, even so late as the reign of Shah Jahan, when his son Dara Shekoh drew a canal from that place to his hunting seat at Shekohpura, which is also called the Ayak, or Jhilri canal.
The Buddhist notices of Sakala refer chiefly to its history in connection with Buddhism. There is the legend of the seven kings who went towards Sagal to carry off Prabhavati, the wife of king Kusa.1 But the king, mounting an elephant, met them outside the city, and cried out with so loud a voice, " I am Kusa," that the exclamation was heard over the whole world, and the seven kings fled away in terror. This legend may have some reference to the seven brothers and sisters of Amba-Kāpa, which is only 40 miles to the east of Sangala. Before the beginning of the Christian era Sagal was the capital of Raja Milinda, whose name is still famous in all Buddhist countries as the skilful opponent of the holy Nagasena.2 The territory was then called Yona, or Yavana, which might refer either to the Greek conquerors, or to their Indo-Scythian successors ; but as Nagasena is said to have lived either 400 or 500 years after Buddha, the date of Milinda is uncertain. Milinda himself states that he was born at Alasadda, which was 200 yojans, or about 1400 miles, distant from Sagal. He was therefore undoubtedly a foreigner ; and, in spite of the
1 Hardy, ' Manual of Buddhism,' 2G3, note. 2 Ibid., 513.
[p.187]: exaggerated distance, I would identify Ms birthplace with. Alexandria Opiane, at the foot of the Indian Caucasus, about 40 miles to the north of Kabul. At a somewhat later period, Sakala was subject to Mahirkul, or Mihirkul, who lost his kingdom by an unsuccessful campaign against Baladitya, king of Magadha. But being afterwards set at liberty by the conqueror, he obtained possession of Kashmir by treachery. I know of no other mention of Sakala until A.D. 633, when it was visited by Hwen Thsang, who describes the neighbouring town of Tse-kia as the capital of a large kingdom, which extended from the Indus to the Byas, and from the foot of the hills to the confluence of the five rivers.
The classical notices of Sangala are confined to the two historical accounts of Arrian and Curtius, and a passing mention by Diodorus. Curtius simply calls it " a great city defended not only by a wall, but by a swamp (palus)"1 But the swamp was a deep one, as some of the inhabitants afterwards escaped by swimming across it (paludem transnavere). Arrian calls it a lake, λίμνη but adds that it was not deep, that it was near the city wall, and that one of the gates opened upon it. He describes the city itself as strong both by art and nature, being defended by brick walls and covered by the lake. Outside the city there was a low hill, Ύήλοφος, which the Kathaeans had surrounded with a triple line of carts for the protection of their camp.2 This little hill I would identify
1 Vita Alex., ix. 1: "Ad magnam deinde urbem pervenit, non muro solum, sed etiam palude munitam."
2 ' Anabasis,' v. 22 : <greek>
[p.188]: with the low ridge to the north-west, called Mundapapura, which would certainly appear to have been outside the city walls, as the broken bricks and pottery do not extend so far.1 I conclude that the camp on the hill was formed chiefly by the fugitives from other places, for whom there was no room in the already crowded city. The hill must have been close to the city walls, because the Kathaeans, after the second line of carts had been broken by the Greeks, fled into the city and shut the gates. It is clear, therefore, that the triple row of carts could only have surrounded the hill on three sides, and that the fourth side was open to the city. The hill was thus connected with the city as a temporary out-work, from which the defenders, if overpowered, could make their escape behind the walls. As the number of carts captured by Alexander was only 300, the hill must have been a very small one ; for if we allow 100 carts to each line, the innermost line, where they were closely packed, at 10 feet per cart, could not have been more than 1000 feet in length round the three sides at the base. Placing the middle row 50 feet beyond the inner one, its length would have been 1200 feet, and that of the outer row, at the same distance, would have been 1400 feet, or little more than a quarter of a mile. Now this accords so well with the size of the Mundapapura hill, that I feel considerable confidence in the accuracy of my identification. As these carts were afterwards used by Ptolemy to form a single line of barrier outside the lake, we obtain a limit to its size, as 300 carts would not have extended more than 0000 feet, or about 17 feet per cart, if placed cud to
1 See Map No. VIII.
[p.189]: end ; but as there may have been numerous trees on the bank of the lake, the length of the barrier may be extended to about 6000 feet. Now it is remarkable that this is the exact length of this outer line according to my survey, which shows the utmost extent of the lake in the rainy season. I could find no trace of the rampart and ditch with which Alexander surrounded the town, but I was not disappointed, as the rains of two thousand years must have obliterated them long ago.
The Kathaeans made an unsuccessful attempt to escape across the lake during the night, but they were checked by the barrier of carts, and driven back into the city. The walls were then breached by undermining, and the place was taken by assault, in which the Kathaeans, according to Arrian, lost 17,000 slain, and 70,000 prisoners. Curtius, however, gives the loss of the Kathaeans at 8000 killed. I am satisfied that Arrian's numbers are erroneous, either through error or exaggeration, as the city was a small one, and could not, at the ordinary rate of 400 or 500 square feet to each person, have contained more than 12,000 people. If we double or triple this for the influx of fugitives, the whole number would be about 30,000 persons. I should like, therefore, to read Arrian's numbers as 7000 slain and 17,000 prisoners. This would bring his number of slain into accord with Curtius, and his total number into accord with probability.
Both Curtius and Arrian agree in stating that Alexander had crossed the Hydraotes before he advanced against Sangala, which should therefore be to the east of that river. But the detailed measurements of
[p.190]: Hwen Thsang are too precise, the statement of the Mahabharata is too clear, and the coincidence of name is too exact to be set aside lightly. Now, the accounts of both Arrian and Curtius show that Alexander was in full march for the Ganges when he heard " that certain free Indians and Kathaeans were resolved to give him battle if he attempted to lead his army thither." Alexander no sooner heard this than he immediately directed his march against the Kathaeans, that is, he changed the previous direction of his march, and proceeded towards Sangala. This was the uniform plan on which he acted during his campaign in Asia, to leave no enemy behind him. When he was in full march for Persia, he turned aside to besiege Tyre ; when he was in hot pursuit of Bessus, the murderer of Darius, he turned to the south to subdue Drangiana and Arachosia ; and when he was longing to enter India, he deviated from his direct march to besiege Aornos. With the Kathaeans the provocation was the same. Like the Tyrians, the Drangians, and the Bazarians of Aornos, they wished to avoid rather than to oppose Alexander ; but if attacked they were resolved to resist. Alexander was then on the eastern bank of the Hydraotcs, or Ravi, and on the day after his departure from the river he came to the city of Pimprama, where he halted to refresh his soldiers, and on the third day reached Sangala. As he was obliged to halt after his first two marches, they must have been forced ones, of not less than 25 miles each, and his last may have been a common march of 12 or 15 miles. Sangala, therefore, must have been about 60 or 65 miles from the camp on the bank of the Hydraotcs. Now this is the exact distance of the Sangala
[p.191]: hill from Lahor which was most probably the position of Alexander's camp when he heard of the recusancy of the Kathaei. I believe, therefore, that Alexander at once gave up his march to the Ganges, and re-crossed the Ravi to punish the people of Sangala for daring to withhold their submission.
Taki or Asarur
[p.191]: I have already mentioned Asarur as the probable position of Hwen Thsang's Tse-kia (Taka), which was the capital of the Panjab in A.D. 633. It is situated about 2 miles to the south of the high-road between Lahor and Pindi Bhatiyan, being 45 miles from the former, and 24 from the latter place.1 It is 19 miles distant from Sangala by the road, but not more than 16 miles in a direct line across the country. Nothing whatever is known of its ancient history, but the people say that it was originally called Udamnagar, or Uda-Nagari, and that it was deserted for many centuries, until Akbar's time, when Ugar Shah, a Dogar, built the Masjid, which still exists, on the top of the mound. The antiquity claimed for the place is confirmed by the large size of the bricks, 18 by 10 by 3 inches, which are found all over the ruins, and by the great number of Indo-Scythian coins that are dis- covered annually after heavy rain. It therefore reaches back to the first century before the Christian era, and from its position I believe it to be the Pimprama of Alexander.
The ruins of Asarur consist of an extensive mound 15,600 feet, or nearly 3 miles in circuit. The highest point is in the north-west quarter, where the mound
1 See Map No. VI.
[p.192]: rises to 59 feet above the fields. This part, which I take to be the ancient palace, is 600 feet long and 400 feet broad, and quite regular in shape. It contains an old well 21 feet in diameter, which has not been used for many years, and is now dry. The palace is completely surrounded by a line of large mounds about 25 feet in height, and 8100 feet, or 1½ mile in circuit, which was evidently the stronghold or citadel of the place. The mounds are rounded and prominent, like the ruins of large towers or bastions. On the east and south sides of the citadel the mass of ruins sinks to 1 and 15 feet in height, but it is twice the size of the citadel, and is, no doubt, the remains of the old city. I could find no trace of any ancient buildings, as all the surface bricks have been long ago carried off to the neighbouring shrine of Ugar Shah at Khangah Masrur ; but amongst the old bricks forming the surrounding wall of the Masjid I found three moulded in different patterns, which could only have belonged to buildings of some importance. I found also a wedge-shaped brick 15 inches long and 3 inches thick, with a breadth of 10 inches at the narrow end, and nearly 10½ inches at the broad end. This could only have been made for a stupa, or a well, but most probably for the latter, as the existing well is 21 feet in diameter. Asarur is now a small village of only 45 houses.
Hwen Thsang places Tse-kia at 14 or 15 li, or 2½ miles, to the north-east of Sakala ; but as there are no traces of any former town in this position, I think it very probable that the true numbers should be 114 or 115 li, or 19 miles, which is just the distance between Sangala and Asarur by the road, although in a direct
[p.193]: line it is not more than 16miles. The circuit of Tse-kia was about 20 li, or upwards of three miles, which agrees sufficiently well with my measurement of the ruins of Asarur at 15,600 feet, or just three miles. At the time of Hwen Thsang's visit there were ten monasteries, but very few Buddhists, and the mass of the people worshipped the Brahmanical gods. To the north-east of the town at 10 li, or nearly 2 miles, there was a stupa of Asoka, 200 feet in height, which marked the spot where Buddha had halted, and which was said to contain a large quantity of his relics. This stupa may, I think, be identified with the little mound of Sālār, near Thata Syadon, just two miles to the north of Asarur.