Chapter XIII: Part 13

Herodotos was the first writer who communicated to the Western nations the story of these ants. He relates it thus (iii. 102): “There are other Indians bordering on the city of Kaspatyros and the country of Paktyike (Afghânistân) settled northward of the other Indians, who resemble the Baktrians in the way they live. They are the most warlike of the Indians, and are the men whom they send to procure the gold (paid in tribute to the King of Persia), for their country adjoins the desert of sand. In this desert then and in the sand there are ants, in size not quite so big as dogs, but larger than foxes. Some that were captured were taken thence, and are with the King of the Persians. These ants, forming their dwelling underground, heap up the sand as the ants in Greece do, and in the same manner; and are very like them in shape. The sand which they cast up is mixed with gold. The Indians therefore go to the desert to get this sand, each man having three camels ... (c. 105). When the Indians arrive at the spot they fill their sacks with the sand, and return home with all possible speed. For the ants, as the Persians say, having readily discovered them by the smell, pursue them, and, as they are the swiftest of all animals, not one of the Indians could escape except by getting the start while the ants were assembling.”

Nearchos (quoted by Strabo, xv. 705) says that he saw skins of the ants which dig up gold as large as the skins of leopards. Megasthenes also (as quoted in the same passage) says that among the Dardai, a populous nation of the Indians living towards the east and among the mountains, there was a mountain plain of about 3000 stadia in circumference, below which were mines containing gold, which ants not less in size than foxes dig up. In winter they dig holes and pile up the earth in heaps, like moles, at the pit-mouths. Pliny (xi. 31) repeats the story in these terms: “The horns of the Indian ant fixed up in the temple of Hercules at Erythrae were objects of great wonderment. These ants excavate gold from mines found in the country of those Northern Indians who are called the Dardae. They are of the colour of cats and of the size of Egyptian wolves. The Indians steal the gold which they dig up in winter during the hot season when the ants keep within their burrows to escape the stifling sultriness of the weather. The ants, however, aroused by the smell, sally out and frequently overtake and mangle the robbers, though they have the swiftest of camels to aid their flight.” It is now understood that the gold-digging ants were neither, as the ancients supposed, an extraordinary kind of real ants, nor, as many learned men have since supposed, larger animals mistaken for ants, but Tibetan miners who, like their descendants of the present day, preferred working their mines in winter when the frozen soil stands well and is not likely to trouble them by falling in. The Sanskrit word pîpilika denotes both an _ant_ and a particular kind of _gold_.

The Dards consist now of several wild and predatory tribes which are settled on the north-west frontier of Kashmir and by the banks of the Indus. The gryphons who guarded the gold were Tibetan mastiffs, a breed of unmatched ferocity. Gold is still found in these regions.

NOTE I.—TAXILA

Pliny, in his _Natural History_ (vi. 21), gives sixty miles as the distance from Peukolatis (Hashtnagar) to Taxila. This would fix its site somewhere on the Haro river to the west of Hasan Abdâl, or just two days’ march from the Indus. But according to the itineraries of the Chinese pilgrims, Fa-Hian and Hwen Thsiang, Taxila lay at three days’ journey to the east of the Indus, and as they made that journey, their authority on the point cannot be questioned. Taxila, it may be therefore concluded, must have been situated in the immediate neighbourhood of Kâla-ka-Sarâi. Now at the distance of just one mile from this place, near the rock-seated village of Shah-Dheri, Cunningham discovered the ruins of a fortified city scattered over a wide space, extending about three miles from north to south, and two miles from east to west, and these ruins he took to be those of Taxila. They lie about eight miles south-east of Hasan Abdâl, thirty-four miles west from the famous tope of Manikyâla, and twenty-four miles north-west from Rawal Pindi. The most ancient part of these ruins, according to the belief of the natives, is a great mound rising to a height of sixty-eight feet above the bed of the stream, called the Tabrâ Nala, which flows past its east side. Cunningham’s identification has now been accepted by all archaeologists, and a Greek text hitherto neglected strikingly confirms its correctness. This text is to be found in the Pseudo-Kallisthenes, and I here translate the remarks made upon it by Sylvain Lévi in a paper which he submitted last year to the Société Asiatique, and which will be found printed at pp. 236, 237 in the 15th volume of the 8th series of the _Journal_ of that society: “The Pseudo-Kallisthenes dwells complacently on the sojourn of Alexander at Taxila and his conversations with the Brahmans. The Brahmans (III. xii. 9, 10) blame the conduct of Kalanos, who, in violation of the duties of his caste, went to live with the Macedonians. ‘It has not pleased him,’ say they, ‘to drink the water of wisdom at the river Tiberoboam.’ And further on (III. xiii. 12) they ask, ‘How could Alexander be the master of all the world when he has not yet gone beyond the river Tiberoboam?’ The Latin of Julius Valerius gives, in the first case, Tiberunco fluvio; in the second, Tyberoboam. The various readings of the Greek manuscripts, indicated by C. Müller in his edition (Didot, 1846), give Boroam, Baroam, Tiberio-potamos, and lastly (MS. A.) Tibernabon. The site fixed by Cunningham for the city of Taxila is distinctly traversed by a river called Tabrâ Nala, which divides into two the ancient city, and washes the foot of the citadel. The ease of confounding the β with λ in the manuscripts permits the correction of Tibernabon into Tibernalon. The essential part of the name is, moreover, Tabrâ, _nala_ being a designation common to small affluents. The resemblance of the two words Tabrânala and Tibernalos is at once apparent; the persistence of geographical names has nothing surprising in it, especially in India. The city of Takshaśila ought then to be placed definitely on the banks of the Tabrânala (a small affluent of the Haro, which bends its course to the Indus, into which it falls twelve miles below Attock) in the position proposed by General Cunningham.”

Taxila, as Alexander found it, was very populous, and possessed of almost incredible wealth. Pliny states that it was situated on a level where the hills sink down into the plain, while Strabo praises the soil as extremely fertile from the number of its springs and water-courses. The Chinese pilgrim, Hwen Thsiang, by whom it was visited in 630 A.D., and afterwards in 643, confirms what Strabo has reported. Taxila, which in Ptolemy’s _Geography_ appears as _Taxiala_, represents either the Sanskrit Takshaśilâ, _i.e._ “hewn stone,” or, more probably, Takshakaśilâ, _i.e._ “Rock of Takshaka,” the great Nâga King. Others, however, take it to represent the Pali Takkasila, _i.e._ the rock of the Takkas, a powerful tribe which anciently occupied the regions between the Indus and the Chenâb (v. _Journal of the Royal Asiatic Society_, vol. xx. p. 343). The famous Aśôka, the grandson of Chandragupta (Sandrokottos), resided in Taxila during the lifetime of his father, Vindusâra, as viceroy of the Panjâb. About the beginning of the second century B.C. Taxila appears to have formed part of the dominions of the Graeco-Baktrian king, Eukratides. In 126 B.C. it was wrested from the Greeks by the Sus or Abars, with whom it remained for about three-quarters of a century, when it was conquered by the Kushân tribe under the great Kanishka. In the year 42 A.D. it is said to have been visited by Apollonios of Tyana and his companion, the Assyrian Damis, who wrote a narrative of the journey, which Philostratos professes to have followed in his life of Apollonios. In 400 A.D. it was visited by Fa-Hian, who calls it Chu-sha-shi-lo, _i.e._ “the severed head,” the usual name by which Taxila was known to the Buddhists of India (_v._ Cunningham’s _Anc. Geog. of India_, pp. 104-121).

NOTE J.—SITE OF ALEXANDER’S CAMP ON THE HYDASPÊS

It is a point of great importance to determine where the camp was situated which Alexander formed on reaching the Hydaspês, and which he made his headquarters till he effected the passage of the river. Without knowing this it cannot with certainty be decided at what point he made the passage or where he defeated Pôros, or where he founded the two cities, Nikaia and Boukephala, which he built after his victory. Such high authorities as Sir A. Burnes, General Court, and General Abbott have placed the camp at Jhîlam, but Lord Elphinstone and General Cunningham prefer Jalâlpûr, a place some thirty miles lower down the stream. These writers all drew their conclusions from personal knowledge of the localities concerned. Cunningham, who wrote later than the others, visited the Hydaspês in 1863, and in his _Geography of Ancient India_ (pp. 157-179) gives an account of the scope and results of his investigations. He points out that Alexander in advancing from Taxila to the Hydaspês had two roads—one called the _upper_, which proceeded through a rich and fertile country, past Rawal Pindi, Manikyâla, and Rohtâs to Jhîlam, and another called the _lower_, which proceeded, with an inclination to southward, to Dudhiâl, and thence by Asanot and Vang to Jalâlpûr. He then shows from Strabo and Pliny that Alexander must have advanced by the _lower_ road. According to Strabo (XV. i. 32), “the direction of Alexander’s march, as far as the Hydaspês, was, for the most part, towards the south; after that, to the Hypanis, it was more towards the _east_.” Now, if Alexander had taken the route by Jhîlam he would have advanced in one continuous straight line, which is in direct opposition to the explicit statement of Strabo, which makes him deviate towards the south. Pliny again (vi. 21), quoting from Diognêtos and Baitôn, the _mensores_ of Alexander, gives the distance from Taxila to the Hydaspês as 120 (Roman) miles. In comparing this distance with that from Shah-Dheri to Jhîlam and Jalâlpûr respectively, we must reject Jhîlam, which is no less than sixteen miles short of the recorded distance, while Jalâlpûr differs from it by less than two miles. The same author thinks that the camp probably extended for about six miles along the bank of the river from Shah Kabir, two miles to the north-east of Jalâlpûr, down to Syadpûr, four miles to the west-south-west. In this position the left flank of the camp would have been only six miles from the wooded promontory of Kotera, where he intended to steal his passage across the river. The breadth of the Hydaspês at Jalâlpûr is about a mile and a quarter.

NOTE K.—BATTLE WITH PÔROS

To the accounts of this memorable battle given by Arrian and the four other writers translated in this volume, I here add the account of it given by Polyainos in his work _On the Stratagems of War_ (II. ix. 22):

“Alexander, in his Indian expedition, advanced to the Hydaspês with intention to cross it, when Porus appeared with his army on the other side determined to dispute his passage. Alexander then marched towards the head of the river, and attempted to cross it there. Thither also Porus marched, and drew up his army on the opposite side. He then made the same effort lower down; there, too, Porus opposed him. Those frequent appearances of intention to cross it, without ever making one real attempt to effect it, the Indians ridiculed, and, concluding that he had no real design to pass the river, they became more negligent in attending his motions, when Alexander, by a rapid march gaining the banks, effected his purpose on barges, boats, and hides stuffed with straw, before the enemy had time to come up with him, who, deceived by so many feint attempts, yielded him at last an uninterrupted passage. In the battle against Porus, Alexander posted part of his cavalry in the right wing, and part he left as a body of reserve at a small distance on the plain. His left wing consisted of the phalanx and his elephants. Porus ordered his elephants to be formed against him, himself taking his station on an elephant at the head of his left wing. The elephants were drawn up within fifty yards of each other, and in those interstices was posted his infantry, so that his front exhibited the appearance of a great wall; the elephants looked like so many towers, and the infantry like the parapet between them. Alexander directed his infantry to attack the enemy in front, while himself at the head of the horse advanced against the cavalry. Against those movements Porus ably guarded. But the beasts could not be kept in their ranks, and, wherever they deserted them, the Macedonians in a compact body pouring in closed with the enemy, and attacked them both in front and flank. The body of reserve, in the meantime wheeling round and attacking their rear, completed the defeat” (_Shepherd’s Translation_).

Grote, referring to this battle, remarks that “the day on which it was fought was the greatest day of Alexander’s life, if we take together the splendour and difficulty of the military achievement and the generous treatment of his conquered opponent.” Military critics cannot point to a single strategical error in the whole series of operations conducted by Alexander himself, or his generals acting under his orders, from the time he encamped on the bank of the Hydaspês till the overthrow and surrender of Pôros. At the same time the courage and skill with which the Indian king contended against the greatest soldier of antiquity, if not of all time, are worthy of the highest admiration, and present a striking contrast to the incompetent generalship and pusillanimity of Darius. “The Greeks,” says General Chesney, “were loud in praises of the Indians; never in all their eight years of constant warfare had they met with such skilled and gallant soldiers, who, moreover, surpassed in stature and bearing all the other races of Asia.... The Indian village community flourished even at that distant period, and in the brave and manly race which fought so stoutly under Porus twenty-two centuries ago we may recognise all the fine qualities of the Punjabi agrarian people of the present day, the gallant men who fought us in our turn so stubbornly, now the most valuable component of the Indian empire, and the best soldiers of its Queen-Empress.”

NOTE L.—THE KATHAIANS

The Kathaioi, it would appear from the text, inhabited the regions lying to the east of the Hydraôtês. Some writers, however, as Strabo informs us (XV. i. 30), placed their country in the tract between the Hydaspês and Akesinês, but this view is manifestly wrong. They are described by ancient authors as one of the most powerful nations of India. Their very name indicates their warlike propensities and predominance, for if it is not identical with that of the military caste, _Kshatriya_, it is at least a modified form of that word. Arrian subsequently (vi. 15) mentions a tribe of independent Indians whose name is a still closer transliteration of _Kshatriya_, the _Xathroi_, whose territories lay between the Indus and the lower course of the Akesinês. Strabo (XV. i. 30) notices some of the peculiar manners and customs of the Kathaians, such as infanticide, and Sati. Lassen has pointed out that their name is connected with that of the Kattia, a nomadic race scattered at intervals through the plains of the Panjâb, but supposed to be the aborigines of the country and of Kolarian descent. Their name occurs in that of the province of Kâthiawâr, which now comprises the province of Gujerat.

NOTE M.—SANGALA

Sir E. H. Bunbury, referring to the uncertainty of the identifications of the tribes and cities of the Panjâb mentioned by Alexander’s historians, says: “While the general course of his march must have followed approximately the same line of route that has been frequented in all ages from the banks of the Indus to those of the Beas, his expeditions against the various warlike tribes that refused submission to his arms led him into frequent excursions to the right and left of his main direction. And with regard to these localities we have a general clue to guide us. The most important of these sites to determine would be that of Sangala, the capital of the Cathaeans, which, according to the narrative of Arrian, was situated between the Hydraotes and the Hyphasis. Hence it was placed by Burnes at Lahore, and by others at Umritsir. But on the other hand there are not wanting strong reasons for identifying Sangala with the Sakala of Indian writers, and this was certainly situated to the west of the Hydraotes, between that river and the Acesines” (_Hist. of Anc. Geog._ pp. 444, 445). This was the view of General Cunningham, who, taking _Śâkala_ or _Sâkala_ (the _Sagala_ of Ptolemy’s _Geography_) to be the name in Sanskrit of the place which the Greeks called _Sangala_, found a site for it at Sânglawâla-Tiba, a small rocky hill with ruins upon it and with a large swamp at its base, and situated between the Râvi (Hydraôtês) and the Chenâb (Akesinês) at a distance of about sixty miles to the west of Lahore. This was no doubt the site of the Śâkala of Sanskrit writers and of the She-kie-lo of the Chinese pilgrim Hwen Thsiang, who visited the place in 630 A.D. But it cannot have been the site of the Sangala of the Greeks, for, in the first place, according to the testimony of all the historians, that city lay between the Hydraôtês and the Hyphasis, and was attacked by Alexander after he had crossed the former river. To meet this objection Cunningham assumes that Alexander must have recrossed the Hydraôtês on hearing that the Kathaians had risen in his rear. His thus turning aside from the direction of his march to make his rear secure is quite consistent with his usual practice, but the historians say nothing from which it can possibly be inferred that on this occasion he made any retrograde movement. But again philology as well as history is adverse to this identification, for, as has lately been shown by M. Sylvain Lévi (_Journal Asiatique_, series viii. vol. xv. pp. 237-239), _Sangala_, in accordance with the rules of transcription, must be taken to represent not _Sâkala_, but _Sâmkala_. Now, just as in Diodôros and Curtius we find Sangala mentioned in connection with a king called Sôphytês, so in an Appendix to Pânini’s _Grammar_, called the Gana-pâtha, _Sâmkala_ is mentioned in connection with _Saubhuta_, which, in accordance with the rules of transcription and the Greek practice of designating Indian rulers after their territories, is evidently the name of the country over which Sôphytês ruled. This country, which was rich and prosperous, as its very name implies, lay between the Hydraôtês and the Hyphasis, probably in the district of Amritsar and towards the hills. Arrian in his narrative of the campaign between these two rivers makes no mention of Sôphytês, or, as he calls him, Sôpeithes, but he afterwards refers to a king of this name whose dominions lay between the Hydaspês and Akesinês. Strabo was aware of the discrepancy of the accounts as to where the dominions of the Kathaians and King Sôphytês were situated.

NOTE N.—ALEXANDER’S ALTARS ON THE HYPHASIS

These altars are mentioned by Pliny, who says (vi. 21): “The Hyphasis was the limit of the marches of Alexander, who, however, crossed it, and dedicated altars on the further bank.” Pliny stands alone in placing these altars on the left bank of the river. The historians all place them on the right bank. Philostratos states that Apollonios of Tyana on his journey into India in the second century of our aera, found the altars still subsisting and their inscriptions still legible. Plutarch affirms that in his days they were held in much veneration by the Praisians, whose kings, he says, were in the habit of crossing the Ganges every year to offer sacrifices in the Grecian manner upon them. It would, however, be unsafe to place much credit in either of these statements. The altars have been sought for in recent times, but not the slightest vestige of them has been discovered. Masson and some other modern writers place them on the Gharra (the united stream of the Vipaśâ and the Śatadru (or Satlej), but this view, while otherwise exposed to serious objections, is upset by the fact alone that in ancient times the two rivers united at a point forty miles below their present junction. As Pliny (vi. 21) gives the distance from the Hyphasis to the Hesidrus (Satlej) along the Royal Road at 169 miles, it is evident that the altars must have been situated at a point high above the junction of these two rivers. V. de Saint-Martin is inclined to think that the altars may have been situated near a chain of heights met with in ascending the Beiâs, and known locally under the name of the _Sekandar-giri_, that is, “Alexander’s mountain.” These heights are at no great distance from Râjagiri, a small and obscure place, but supposed to represent Râjagriha mentioned in the _Râmâyana_ as the capital of a line of princes called the Aśvapati (or _Assapati_ in Prakrit) who governed the Kekaya, or, as Arrian calls them in his _Indika_, the Kêkeoi. Lassen, followed by Saint-Martin, identified Sôpeithes as belonging to the line of princes indicated. The identification has been superseded by a better, but Saint-Martin’s argument, as far as it concerns the position of the altars, is not thereby affected. Sir E. H. Bunbury considers that the point where Alexander erected the twelve altars cannot be regarded as determined within even approximate limits. It appears probable, he thinks, that they were situated at some distance above the confluence of the two rivers, and not very far from the point where the Beas emerges from the mountain ranges. We learn indeed, he adds, that throughout his advance Alexander kept as near as he could to the mountains; partly from the idea that he would thus find the great rivers more easily passable, as being nearer their sources; partly from an exaggerated impression of the sterile and desert character of the plains farther south (_Hist. of Anc. Geog._ p. 444).

NOTE O.—VOYAGE DOWN THE HYDASPÊS AND AKESINÊS TO THE INDUS

From the point of embarkation at Nikaia (Mong) to the confluence of the united streams of the Panjâb with the Indus, the distance in a straight line may be reckoned at about 300 miles. Alexander in descending to this confluence had no sooner left the dominions of Pôros than he was engaged in a constant succession of hostilities with the riparian tribes. He had no intention of leaving India as a fugitive. He must depart as a conqueror and master of all wherever he appeared. He had no wish, therefore, even had it been possible, to drop quietly down stream to the ocean. He must demand submission to his authority from all the tribes he might encounter on his way, and, if this were refused, enforce it at the sword’s point. These tribes were the bravest of the brave in India—the very ancestors of the Rajputras, or Rajputs, whose splendid military qualities have spread their fame throughout the world. Such of these tribes as inhabited the fertile regions adjacent to the rivers seemed to have settled in towns and villages and to have practised agriculture, while those that tenanted the deserts which extended far eastward into the interior led a half-wandering pastoral life, and subsisted as much on the produce of rapine as on the produce of their flocks and herds. They were all proudly jealous of their independence, and owned no authority but that of their proper chiefs. Though they were separated into distinct tribes, which were almost perpetually at feud, they were still able when confronted with a common danger to combine into formidable confederacies. In all times they have opposed to invasion a vigorous and sometimes a desperate resistance (_v._ Saint-Martin, _Étude_, p. 113).

NOTE P.—THE MALLOI AND OXYDRAKAI

The names of these two warlike tribes are very frequently conjoined in the narratives of the historians. In Sanskrit works they appear as the Mâlava and the Kshudraka, and a verse of the _Mahâbhârata_ combines them in a single appellation, _Kshudrakamâlava_. They are mentioned in combination by Pânini also as two Bahîka people of the north-west. Arrian (_Indika_, c. iv.) places the Oxydrakai on the Hydaspês _above_ its confluence with the Akesinês. It is doubtful, however, that this was their real position. Bunbury inclines to think that they lay on the east or left bank of the Satlej—the province of Bahawalpur—and that they may very well have extended as far as the junction of the Satlej with the Indus and the neighbourhood of Uchh. General Cunningham, he adds, is alone in placing the Oxydracae to the north of the Malli. That author has, however, the _Indika_ to support his view. Their name in the classics appears in various forms, Strabo and Stephanos Byz. calling them _Hydrakai_, Pliny _Sydracae_, and Diodôros _Syrakousai_. Strabo says they were reported to be the descendants of Bacchos because the vine grew in their country, and because their kings displayed great pomp in setting out on their warlike expeditions after the Bacchic manner (XV. i. 8). They are no doubt to be identified with the Śudras, whose name in early times did not denote a caste, as it did afterwards, but a tribe of aborigines, or, at all events, a tribe of non-Aryan origin. The final _ka_ in the Greek form of their name is a common Sanskrit suffix to ethnic names given or withheld at random. The single combat between Dioxippos and a Macedonian bravo called Horratas took place after a great banquet at which Alexander entertained the envoys of the Oxydrakai.

The territory of the Malloi was of great extent, comprehending a part of the Doâb formed by the Akesinês and the Hydraôtês, and extending, according to Arrian (_Indika_, c. iv.), to the confluence of the Akesinês and the Indus. In the _Mahâbhârata_ they figure as a great people, being there distinguished into the Eastern, Southern, and Western Mâlavas (_Mahâbh._ vi. 107). They are mentioned also in the inscription of Samudragupta (of the first half of the third century A.D.) among other peoples of the Panjâb who were subject to the King of Madhya-desha (_v._ V. de Saint-Martin, _Étude_, pp. 116-120). “These two races,” says Thirlwall (_History of Greece_, vii. 40), “were composed of widely different elements; for the name of one appears to have been derived from that of the Sudra caste; and it is certain that the Brahmins were predominant in the other. We can easily understand why they did not intermarry and were seldom at peace with each other.” The feud, however, may have been one of race rather than of caste, though no doubt the distinctions of caste originated in difference of race.

NOTE Q.—THE CAPITAL OF THE MALLOI

Diodôros and Curtius assign this city to the Oxydrakai, but erroneously. General Cunningham identifies it with Multân and takes it to be also the capital of the Malloi “to which many men from other cities had fled for safety.” Arrian seems, however, to indicate that the two places were distinct. V. de Saint-Martin inclines to identify the Mallian capital with Harrâpa (the Harapa which Cunningham takes to be the city captured by Perdikkas). Multân is at present the capital of the province of the same name, which comprises pretty nearly the same territories as those occupied by the Malloi of the Greek historians. Multân is not situated on the Râvi now, but on the Chenâb, and at a distance of more than thirty miles _below_ the junction of that river with the Râvi. This circumstance would be quite fatal to Cunningham’s view if the junction had not shifted. But it has shifted, for in Alexander’s time the rivers met about fifteen miles below Multân. “The old channel (Cunningham says) still exists and is duly entered in the large maps of the Multân division. It leaves the present bed at Sarai Siddhû and follows a winding course for thirty miles to the south-south-west, when it suddenly turns to the west for eighteen miles as far as Multân, and, after completely encircling the fortress, continues its westerly course for five miles below Multân. It then suddenly turns to the south-south-west for ten miles, and is finally lost in the low-lying lands of the bed of the Chenâb. Even to this day the Râvi clings to its ancient channel, and at all high floods the waters of the river still find their way to Multân by the old bed, as I myself have witnessed on two occasions. The date of the change is unknown, but was certainly subsequent to A.D. 713.” From Arrian’s narrative it would appear that Alexander occupied three days, one of which was spent in rest, in advancing from the city of the Brachmans to the city of the Malloi. The distance traversed would be thirty-four miles, if Cunningham’s identification of the former city with Atâri and of the latter with Multân be correct. The city where Alexander was wounded appears from Arrian’s account to have been at some distance from the Hydraôtês, and if so could not have been Multân.

NOTE R.—ALEXANDER IN SINDH

Arrian and the other historians of Alexander have treated very briefly and vaguely his campaign in the valley of the Indus. Hence it is difficult to trace the course of his operations as he descended from the great confluence at Uchh to Patala where the Indus bifurcates to form the Delta. The distance between these two points, if measured by the course of the river, may be estimated at nearly four hundred miles, yet we find, as Saint-Martin observes, that in the descent not a single distance is indicated, nor a single peculiar feature of the country described which might serve as a sign-post for the direction and guidance of our inquiries. It is at the same time difficult to reconcile the discrepancies found to exist in the accounts transmitted to us, and altogether the search for identities must here mainly concern itself with the names of tribes. In determining how these tribes were collocated it is necessary to take cognisance of the changes which have taken place in the course of the Indus since Alexander’s time. Captain M’Murdo was the first to call attention (in 1834) to these changes, which were not confined to the terminal course of the river, but extended more than two hundred miles above the Delta. He proved that up to the seventh century of our aera the main stream of the Indus, instead of following its present channel, pursued a more direct course to the sea some sixty or seventy miles farther east than it now flows. The old channel, which leaves the present stream at some distance above Bhakar, passes the ruins of Alôr, and then proceeds directly towards the south nearly as far as Brâhmanâbâd, above which it divides into two channels, one rejoining the present course above Haidarâbâd, while the other pursues a south-easterly course towards the Ran of Kachh. The voyage down the lower part of the course took place during the season of the inundation when the plains were laid far and wide under water, and the current was rapid and violent. As the march followed mainly the line of the river the country would appear to the Macedonians extremely rich, fertile, and populous, while the sterility of the regions that lay beyond the reach of the inundations would seldom be brought under their cognisance. In descending the river they could not fail to notice the contrast presented by the plains on its opposite banks, those on the east exhibiting a uniform expanse without any visible boundary, while those on the west were hemmed in by a great mountain rampart which in running southwards gradually approached the Indus till the roots of the hills were laved by its waters. The inhabitants would strike them as being more swarthy in their complexion than the men of the Panjâb, from whom they differed also in their political predilections, as they preferred kingly government to republican independence, and allowed the Brahman to exercise a decisive influence over public life. The descent of the Indus by Alexander, as Bunbury remarks, may be considered as constituting a kind of aera in the geographical knowledge of the Greeks. It does not appear, he adds, that it was ever repeated; and while subsequent researches added materially to the knowledge possessed by the Greeks of the valley of the Ganges and the more easterly provinces of India, their information concerning the great river Indus and the regions through which it flows continued to be derived almost exclusively from the voyage of Alexander and the accounts transmitted by the contemporary historians.

After leaving the great confluence the first tribe Alexander reached were the _Sogdoi_, who appear as the _Sodrai_ in Diodôros, who states that Alexander founded among them on the banks of the river a city called Alexandreia in which he placed 10,000 inhabitants. The Sogdoi have been identified with the _Sohda_ Rajputs who now occupy the south-east district of Sindh about Amarkot, but who in former times held large possessions on the banks of the Indus to the northward of Alôr. This place, though now only a scene of ruins, was formerly, before it was deserted by the river, one of the largest and most flourishing cities in all Sindh. Saint-Martin takes it to have been the Sogdian capital, and thinks that the city which Alexander founded lay in its vicinity at Rôri, since right opposite to this place there rose in the middle of the river the rocky island of Bhakar, which presented every natural advantage for the site of a great fortress. Cunningham, however, would place the capital higher up stream, about midway between Alôr and Uchh, at a village which appears in old maps under the name of Sirwahi, and which may possibly represent the _Seori_ of Sindh history and the _Sodrai_ of Diodôros. In this neighbourhood lies the most frequented ghât for the crossing of the Indus towards the west _viâ_ Gandâva and the Bolan Pass; and as the ghâts always determine the roads, it was probably at this point of passage Krateros recrossed the Indus when he was despatched with the main body of the army and the elephants to return home through the countries in which that Pass lay. The name _Sodrai_, some think, represents the Sanskrit _Śûdra_ which designates the servile or lowest of the four castes. If this be so, the Sodrai may be regarded as a remnant of the primitive stock which peopled the country before the advent of the Aryans (_v._ Lassen, _Ind. Alt._ ii. p. 174; Saint-Martin, _Étude_, pp. 150-161; Cunningham, _Anc. Geog. of India_, pp. 249-256).

NOTE S.—SINDIMANA

Sambus is called Sabus by Curtius, who, without giving the name of his capital, informs us that Alexander captured it by mining and then marched to rejoin his fleet on the Indus. The Greek name of this capital, _Sindimana_, has led to its identification with Sehwân, a site of very high antiquity. The great mound which was once its citadel has been formed chiefly of ruined buildings accumulated in the course of ages on a scarped rock at the end of the Lakki range of hills. Its water supply is at present entirely derived from the Indus, which not only flows under the eastern front of the town, but also along the northern by a channel from the great Manchur Lake, which perhaps formerly extended up even to the city walls. The objection to this identification, that Sehwân’s position on the Indus conflicts with the statement that Alexander had to march from Sindomana to reach that river, is removed by the fact that the Indus has changed its course since Alexander’s time. Wilson derives the Greek Sindomana from what he calls a very allowable Sanskrit compound, _Sindu-mân_, “the possessor of Sindh.” Cunningham, however, would refer the name to _Saindava-vanam_ or _Sainduwân_, “the abode of the Saindavas.” _v._ his _Anc. Geog. of India_, pp. 263-266, and also Saint-Martin’s _Étude_, p. 166, where it is stated that the name of Sambus is probably connected with that of the tribe called Sammah, whose chiefs have at different epochs played a distinguished part in the valley of the Indus. In Hindu mythology _Samba_ is the son of Krishna. According to Plutarch, it was somewhere in the dominions of Sabbas that Alexander had his interview with the ten Indian gymnosophists. Sehwân is the Sewistan of the Arabs. According to Burton (writing in 1857), it is a hot, filthy, and most unwholesome place, with a rascally population (of 6000) which includes many beggars and devotees. _v._ his _Sindh_, p. 8. The population has since increased to upwards of 160,000.

NOTE T.—CITY OF THE BRACHMANS—HARMATELIA

This city of the Brachmans Cunningham takes to have been Brâhmana, or Brâhmanâbâd, which was ninety miles distant by water from Marija Dand, the point where he supposes Alexander rejoined his fleet after the capture of Sindimana. Brâhmanâbâd was situated on the old channel of the Indus forty-seven miles to the north-east of Haidarâbâd or Nirankot, the Patala of ancient times. Shortly after the Mohammedan conquest it was supplanted by Mansûra, which either occupied its site or lay very near it, as, according to Ibn Haukal, the place was called in the Sindh language Bâmiwân. It was destroyed by an earthquake sometime before the beginning of the eleventh century. Its ruins were discovered at Bambhra-ka-thûl by Mr. Bellasis, whose excavations have shown conclusively the truth of the popular tradition which ascribed its downfall to an earthquake. Cunningham further thinks that Brâhmanâbâd was the _Harmatelia_ of Diodôros—the place where Ptolemy was wounded by the poisoned arrow. Harmatelia (he says) is only a softer pronunciation of _Brâhma-thala_ or Brahmana-sthala, just as Hermes, the phallic god of the Greeks, is the same as Brahmâ, the original phallic god of the Indians. He thinks that the king whom Justin (xii. 10) calls _Ambiger_ was no other than Mousikanos, whose dominions extended as far south as the Delta, _Ambiger_ being his family name and Mousikanos his dynastic title (_Geog._ pp. 267-269). Saint-Martin, on the other hand, recognises Harmatelia in a place variously designated by Arab writers _Armael_, _Armaïl_, _Armâbil_, and _Armatel_, but of which the position is unknown (_Étude_, pp. 167, 168). In his ancient map of India Colonel Yule, who takes the same view as Saint-Martin, identifies Harmatelia with Bela.

NOTE U.—PATALA

The situation of Patala has been a fertile theme of controversy. Arrian seems, no doubt, to give here a clear indication of its position in saying that it stood near where the Indus bifurcates; but as this point has from time to time shifted, the controversy has turned mainly on the question where this point is to be fixed. The river bifurcates at present at Mottâri, which lies twelve miles above Haidarâbâd, and it has been known to bifurcate a little above, and also a little below Thatha, at Bauna also, and at Trikul. As a matter of fact, these bifurcations no longer exist, except perhaps for a part of the year when the river is in flood and recurs to some of its old channels. It is not then surprising that various identifications have been proposed for Patala. It was placed at Brâhmanâbâd by M’Murdo, Wilson, and Lassen; at Thatha by Rennell, Vincent, Ritter, and the two brothers, James and Sir Alexander Burnes; and at Haidarâbâd, the Nirankot of Arab writers, by Droysen, Benfey, Burton, Saint-Martin, Cunningham, and Bunbury. The arguments in favour of Haidarâbâd seem to be quite conclusive. They will be found stated at length in Saint-Martin’s _Étude_ (pp. 168-191), and Cunningham’s _Geography_ (pp. 279-287). One of the most cogent is that the dimensions of the Delta, as given by the Greek writers, are only justified if the apex of the Delta is taken to have been in Alexander’s time at or near Haidarâbâd. If the apex had then been as high up as Brâhmanâbâd, or as far down as Thatha, the size of the Delta would be as grossly exaggerated in the one case as it would be underrated in the other. The same conclusion is indicated in the information supplied to the late Dr. Wilson of Bombay by the Brahmans of Sehwân, that, according to their local legends, as recorded in their Sanskrit books, Thatha was _Déval_, and Haidarâbâd _Néran_, and more anciently _Patolpuri_. _Patala_ was thought at one time to have been a transcription of the Sanskrit Pâtâla, _the nether world_, into which the sun descends at the end of his day’s journey, and hence THE WEST; but a better etymology is the Sanskrit _potala_, “a station for ships,” from _pôta_, “a vessel.” The name of the Indian Delta was Patalênê. Haidarâbâd stands on a long flat-topped hill, and Patala, if this was its site, must have occupied a commanding position, the advantages of which, alike for strategy and commerce, Alexander would perceive at a glance. The main stream of the Indus now flows to the west of this position. In the second chapter of his _Indika_ Arrian repeats the statement that the Indus enters the ocean by two mouths. Aristoboulos estimated the interval between them at 1000 stadia, but Nearchos at 1800. The interval from the west to the east arm measures at present 125 British miles. The sea-front of the Egyptian Delta with which the Greeks compared that of the Indus Delta is not less than 160 miles. The Prince of Patala was called _Moeris_.

NOTE V.—ALEXANDER’S MARCH THROUGH GEDRÔSIA-PURA

“No traveller,” says Bunbury, referring to the interior of Mekran, “has as yet traversed its length from one end to the other in the direction followed by Alexander. So far as we can judge, he appears to have kept along a kind of plain or valley, which is found to run nearly parallel to the coast between the interior range of the Mushti (or Washati) hills and the lower ragged hills that bound the immediate neighbourhood of the sea-coast. This line of route has been followed in very recent times by Major Ross from Kedj to Bela, and seems to form a natural line of communication, keeping throughout about the required distance (60 or 70 miles) from the coast [the distance required for maintaining communication with the fleet].... This line of march so far as is yet known does not appear to traverse any such frightful deserts _of sand_ as those described by the historians of Alexander. Nor can the site of Pura ... be determined with accuracy. It has been generally identified with Bunpoor (Banpûr), the most important place in Western Beloochistan, or with Pahra, a village in the same neighbourhood; but the resemblance of name is in this case of little value—_poor_ signifying merely a town—while the remoteness of Bunpoor from the sea, and its position to the north of the central chain of mountains, which Alexander must therefore have traversed in order to reach it, present considerable difficulties in the way of this view” (_Hist. of Anc. Geog._ pp. 519-520). Strabo, in his chapter on Ariana, narrates in graphic detail, like Arrian, the sufferings experienced by the Macedonians in passing through Gedrôsia. The summer, he says, was purposely chosen for leaving India, since rains then fall in Gedrôsia, filling the rivers and wells which fail in winter. Alexander kept at the utmost from the sea not more than 500 stadia in order to secure the coast for his fleet. The army was saved by eating dates and the marrow of the palm-tree, but many persons were suffocated by eating unripe dates.

To account for the surprising length of time (60 days) occupied on this march, which could not have exceeded 400 English miles, we must suppose that the troops were obliged to make frequent halts at places where water was procurable. Strabo says that it was found necessary on account of the watering-places to make marches of two, four, and even sometimes of six hundred stadia generally during the night. The land distances, like the sea distances of Nearchos, seem to have been grossly exaggerated. The march of Semiramis through this desert and that of Cyrus seem to be mythical. Alexander’s loss in men during the march must have been exaggerated by the historians, as he brought the bulk of his army with him to Pura.

NOTE W.—INDIAN SAGES

According to Megasthenes the Indian sages were divided into two sects, Brahmans and Sarmans. There was besides a third sect, described as quarrelsome, fond of wrangling, foolish and boastful. The Brahmans, he says, were held in higher esteem than the Sarmans because there was more agreement in their doctrines. Among the Sarmans the Hylobioi (_living in woods_) were held in most honour, and next to them the physicians, who are mendicants and also ascetics, like the class above them and the class below them, which consisted of sorcerers and fortune-tellers. Megasthenes has related at some length the nature of the opinions and practices of all these sects, and Duncker considers that in all essential points his accounts agree with the native authorities, though the view taken may be here and there too favourable, in some points too advanced, in others not sufficiently discriminating. “It is true,” he says, “that the Brahmans and the initiated of the Enlightened (Buddhists), the Śramanas, are confounded in the order of the sages; this is shown by the statement that any one could enter into this order.... In the description of the life of the ascetics and wandering sages, the Brahmans and Bhikshus (mendicants) are again confounded, and if the Greeks tell us that the severe sages of the forest were too proud to go to the court at the request of the king, the statement holds good according to the evidence of the Epos of the Brahmanic saints, and the Sutras of the great teachers among the Buddhists. In the examination of the doctrines of the Indian sages, Megasthenes distinguished the Brahmans and the Buddhists, inasmuch as he opposes the less-honoured sects to the first, and declares the Brahmans to be the most important. From his whole account it is clear that at his date, _i.e._ about the year 300 B.C., the Brahmans had distinctly the upper hand. But, according to him, the Śramanas took the next place to the Brahmans among the less-honoured sects. Among the Buddhists Śramana is the ordinary name for their clergy” (_Hist. of Antiq._ pp. 422-424).

NOTE X.—THE INDIAN MONTH

Curtius apparently means that the Indians mark time, not by taking a month to be the period from full moon to next full moon, but from new moon to full moon. “The year of the Indians (says Duncker) was divided into 12 months of 30 days; the month was divided into two halves of 15 days each, and the day into 30 hours (_muhurta_). In order to bring this year of 360 days into harmony with the natural time, the Brahmans established a quinquennial cycle of 1860 lunar days. Three years had 12 months of 30 lunar days; the third and fifth year of the cycle had 13 months of the same number of days. The Brahmans do not seem to have perceived that by this arrangement the cycle contained almost four days in excess of the astronomical time, and indeed they were not very skilful astronomers” (_v._ his _History of Antiquity_, iv. 283, 284). According to Weber this system of calculating time was borrowed from the Babylonians, but Max Müller and learned Hindus hold it to be indigenous. The Indian name for the half of a lunar month is _paksha_. The half from new moon to full moon was called at first _pûrva_ (fore), and afterwards _śukla_ (bright); the other half was called _apara_ (posterior), and afterwards _krishna_ (dark). Le Clerc concludes his criticism of this passage thus: “Matthaeus Raderus endeavours to explain Curtius as if he designed to demonstrate that one month began and was understood to commence a little after the change to the full moon, and the next, from the time when she began to decrease to the next change. This, indeed, ought to be his meaning; but it is strangely expressed, when he tells us that the moon begins to show herself horned on the sixteenth day, when ’tis evident she does not appear so till about seven days after full moon. But before Raderus, Thomas Lydiat had tried to solve the matter otherways. However, Scaliger, in his _Prolegomena_ to his _Canones Isagogicae_, p. 11, has plainly showed that Lydiat neither understood Curtius nor Curtius the author which he copied from. The ancient Persians counted 15 days to each of their months, and 24 of these months to the solar year, before the introduction of Mohammedism, as John Chardin evidently demonstrates in his _Itinerarium Persicum_, tome xi. p. 14, quarto” (_v._ Rooke’s _Arrian_, p. 12).

NOTE Y.—BATTLE WITH PÔROS

Mr. Heitland has the following note on this passage: “Arrian (v. 16, sec. 2) tells us that Alexander was making a flanking movement (παρήλαυνεν) with the bulk of his cavalry to attack the enemy’s left wing. He then goes on (sec. 3): _Against the right wing he sent Koinos at the head of his own regiment of horse and that of Dêmêtrios, and ordered him, when the barbarians on seeing what a dense mass of cavalry was opposed to them, should be riding along to encounter it, to hang close upon their rear_,[411] a hard passage, it is true, but one which need not be unintelligible to any one who bears in mind that Alexander’s movement was a flanking one, and reads with care the description of his attack in c. 16, sec. 4, and c. 17, sec. 1, 2. The situation is this: Alexander was not himself in position on the right wing, but put Coenus there with some of the cavalry, while he himself with the main body made the flanking movement. This he did with speed, so as to take the Indian horse in flank, before they had time to change their front and meet him. They tried to execute this movement, but had not time; and while they were in the confusion thus brought about, Coenus fell upon what had been their front, but was now their disordered flank. Whether the Indian horse from their right wing was brought over to succour that on their left or not, does not affect the probable position of Coenus. The one difficulty in the way of this explanation is the presence, according to Arrian, 15, sec. 7, of the war-chariots in front of the Indian horse. But it seems easier to suppose that Coenus was able to elude these clumsy adversaries than that Alexander expected him to see from the Macedonian left the right moment for his own charge, and then wheel round the rear of the whole Indian army and execute his orders opportunely. Diodorus, xvii. 88, says: _The Macedonian cavalry began the action, and destroyed nearly all the chariots of the Indians_.[412] If this refers, as I think it does, to the beginning of the main battle, the chief objection is removed” (_Alexander in India_, pp. 122, 123). This explanation is different from that offered by Moberly, as the reader will see by referring to my note on Arrian, p. 104, _n._ 2.

NOTE Z.—INDIAN SERPENTS