The Ancient Geography of India/Singhapura

From Jatland Wiki
Jump to navigation Jump to search
Wikified by:Laxman Burdak, IFS (Retd.)
Back to Index of the Book
The Ancient Geography of India: I.
The Buddhist Period, Including the Campaigns of Alexander, and the Travels of Hwen-Thsang
Sir Alexander Cunningham
Trübner and Company, 1871 - India

4. Singhapura or Ketas

[p.124]: According to Hwen Thsang, the capital of the kingdom of Seng-ho-pu-lo, or Singhapura, was situated at 700 li, or 117 miles, to the south-east of Taxila. The bearing points to Jhelam, near which is the town of Sangohi, which has been noted by M. Vivien de St. Martin as the possible representative of Singhapura. But Sangohi stands on an open plain, instead of on a high mountain of difficult access, as described by the pilgrim. The vicinity of ten pools of limpid water, with surrounding temples and sculptures, points to the holy tanks of Ketaksh, or Khetas, which are still visited by crowds of pilgrims from all parts of India. I think also that the name of Ketas is only a slightly altered form of the Sanskrit Swetavasa, or the " White

[p.125]: Robes," which Hwen Thsang mentions as the title of the chief religious sect then resident near Singhapura. In the western countries, where the compound sw is changed to kh, the name would have been pronounced Khetavasa, or by a slight contraction, Khetas1 The Brahmans of course refer the name to their own religion, and say that the place was called Kataksha, or the " Raining Eyes," because the tears literally rained from Siva's eyes when he heard of the death of his wife Sati. But as their own spelling of the name Ketaksh, which I received from themselves, is at variance with the meaning which they give to it, I am inclined to adopt the etymology that I have already suggested as Sweta-vasa, or the "White Robes," This sect would appear to have belonged to the Swetambara, or " "White-robed" division of the Jains, while another sect at the same place, who are described by Hwen Thsang as going naked, must be the Digambara, or "unclothed" (literally "sky-clad") division of the Jains. Their books also are stated to have been chiefly copied from the Buddhist literature, while the statue of their god resembled that of Buddha himself. From these curious details it seems almost certain that this heretical sect must have been Jains, whose religion has much in common with Buddhism, while their statues are frequently mistaken for those of Buddha.

Ketas is situated on the north side of the Salt Range, at 16 miles from Pind Dadan Khan, and 18 miles from Chakowal, but not more than 85 miles from Shah-dheri, or Taxila. Now the distance of Singhapura from Taxila is given at 700 li, or 117 miles, which is

1 Thus the Sanskrit Saraswati became the Zend Harakhaiti, and the Greek Arakhotos.

[p.126]: certainly too great, as it would place the capital about 30 miles beyond the most distant point of the hills in any direction between the south and east. Singha-pura is described as situated on the top of a high hill of difficult access ; and as the climate is said to be very cold, it is certain that the place must have occupied one of the isolated peaks either of the Salt Range on the south-south-east, or of the Balnath Range on the east-south-east.1 But as there are no clear pools swarming with fish in the Balnath Range, I have little hesitation in identifying the place described by Hwen Thsang with the beautiful limpid pool of Ketas, which has been esteemed holy from time immemorial.

The capital of Singhapura was situated at from 40 to 50 li, or 7 to 8 miles, to the north-west of the sacred tanks ; but I know of no place that corresponds with this bearing and distance. Malot was the capital of the Janjuhas at a very early period ; but its bearing is south-east, and its distance 12 miles. If we might read 4 to 5 li, instead of 40 to 50, the capital might at once be identified with the ruined fort of Kotera, which is situated on a steep hill to the west, about 200 feet in height, that overhangs the town and holy pools of Ketas. This is called the ancient town. It consists of an upper fort, 1200 feet long, by 300 broad, and of a lower fort. 800 feet long, by 450 broad, the circuit of the two being about 3600 feet, or less than three-quarters of a mile. But the whole circuit of Ketas, including the modern town on both banks of the stream, both above and below the fort, is about 2 miles. This is rather smaller than the capital described by Hwen Thsang, which was 14

1 See Maps Nos. V. and VI.

[p.127]: or 15 li, or 2¼ to 2½ miles, in circuit. But as it corresponds in all other material particulars, I think that Ketas has a very good claim to be identified with the capital of Singhapura.

According to Hwen Thsang,1 the district was 3600 li, or 600 miles, in circuit. On the west it was bounded by the Indus, on the north by the southern frontier of Taxila, 120 miles in length, and on the south by the Jhelam and the northern frontier of Taki, or the plains of the Panjab. It cannot therefore have extended much beyond the foot of the Salt Range.

This limit would make the Indus frontier about 60 miles in length, the Jhelam frontier about 50 miles, and the northern and southern frontiers each 120 miles, or altogether 350 miles. The only explanation that occurs to me of the difference between this number and that of Hwen Thsang, is the probability that the ancient kos of the Panjab was the same as the modern one, that is, a short kos of 1-9/32 mile, or 1 mile and 2¼ furlongs, and that the Chinese pilgrim, ignorant of the difference, made his calculations in the common Indian kos of about two miles. This would reduce his numbers by very nearly one-third, and at the same time bring them into close accordance with the actual measurements of our maps. Thus, Hwen Thsang's 3600 li, or 600 miles, for the circuit of Singhapura, would become 400 miles, which is within 50 miles of the actual measurement already given. Great accuracy cannot be expected in these estimates of frontier distances, as the pilgrim had no means of checking the numbers of his informants. With the road distances which he had himself travelled it was different, as

1 Julien's ' Hiouen Thsang,' ii. 162.

[p.128]: he could test them by his own knowledge of the time occupied, as well as by the number of journeys between any two points. In the present instance of Singhapura it is quite certain that the frontier distance is exaggerated, as the boundary of Tsekia, or Taki, is also said to have extended to the Indus, which could not have been the case if the frontier of Singhapura had stretched further to the south than I have placed it.

5. Punacha or Punach

[p.128]: The district of Puan-nu-tso, or Punacha, is placed by Hwen Thsang at 700 li, or 117 miles, to the south-west of Kashmir.1 It is called Punats by the Kashmiris, who have adopted a soft pronunciation of the ch, as in Pir Pantsal for Panchal of the Panjabis. Moorcroftf spells the name Prunch, or Pruntz, according to the Kashmiris. General Court also has Prunch ; but it is called Punje by Wilford's surveyor, Mirza Mogal Beg, and Punch by Vigne, both of whom actually visited the place. Its distance from Kashmir, as measured on the map via Barahmula and Uri is 75 miles, which is equal to about 100 miles of actual road distance.2

Hwen Thsang describes Punach as 2000 li, or 333 miles, in circuit, which is just about twice its actual size. On the west it is bounded by the Jhelam, on the north by the Pir Panchal range, and on the east and south-east by the small state of Rajaori. But these limits, which include the potty state of Kotali, are not more than 170 miles in circuit ; and even if the tract at the source of the Punach river be included, the frontier

1 Julien's ' Hiouen Thsang,' ii. 187.

2 ' Travels,' ii. 298.

3 See Maps Nos. V. and VI.

[p.129]: will not be more than 200 miles in circuit. But as the distances in the mountain districts were most probably estimated by the lengths of the roads, the circuit of the frontier line may be taken as equivalent to about 300 miles in road distance.

In the seventh century Punach was without a king, and subject to Kashmir ; but in later times it had a chief of its own, whose descendants. Shir Jang Khan and Shams Khan, were put to death by Gulab Singh, of Jammu, and this petty sovereignty once more forms part of the kingdom of Kashmir.

6. Rajapura or Rajaori

[p.129]: From Punach, Hwen Thsang proceeded to the south-east for 400 li, or 67 miles, to Ko-lo-she-pu-lo, or Rajapura1 which I long ago identified with the petty chief ship of Rajaori, to the south of Kashmir. The circuit of the district is described as 4000 li, or 667 miles, which is about double the true amount, unless, as is not improbable, the whole of the hill-states as far as the Ravi be included within its boundaries.

From the native chronicle of Kashmir we learn that the petty chiefships of the hills to the south and south-east of the valley were generally subject to Kashmir ; and there is no reason to suppose that they were independent at the time of Hwen Thsang's visit. The district of Rajaori proper is nearly a square of about 40 miles each side, bounded on the north by the Pir Panchal, on the west by Punach, on the south by Bhimbar, and on the east by Rihasi and Aknur.2 By extending its boundary on the east to the Chenab,

1 Julien's ' Hiouen Thsang,' ii. 188.

2 See Maps Nos. V. and VI.

[p.130]: and on the south to the plains, it would include all these petty places ; even then its frontier would not be more than 240 miles, or by road about 320 miles. But if the frontier of these hill-states subject to Kashmir be extended to the Ravi on the east, the circuit would be about 420 miles measured on the map, or not less than 560 miles by road.

Rajapuri is frequently mentioned during the medieval period of Kashmirian history, but chiefly in the eleventh and twelfth centuries, when it was an independent state under its own rajas. In the fifteenth century the Hindu family was dispossessed in favour of a son of the Muhammadan king of Kashmir ; and his descendant was so reduced by Gulab Singh that in 1846 he was glad to accept an estate in the British district of Kangra in exchange for his petty chiefship of Rajaori.

Hill States of the Punjab

As the Chinese pilgrim has noticed so few of the many hill-states of the Panjab, I will here add a brief outline of the information which I have myself been able to collect regarding them.

According to popular opinion the petty states of the Alpine Panjab, at the present time, consist of twenty-two Muhammadan and twenty-two Hindu chiefships, the former lying to the west, and the latter to the east of the Chenab river. An older classification divides them into three groups, each named after the most powerful state which formed the head of the confederation. These were Kashmir, Dogra, and Trigartta. The first consisted of the rich valley of Kashmir, and all the petty states between the Indus and Jehlam; the second included Jammu, and the other

[p.131]: petty states between the Jehlam and the Ravi ; the third comprised Jalandhar, and the various small states between the Ravi and the Satlej.

This division into three groups most probably existed prior to the seventh century, as we find that the states to the east of the Ravi were quite independent of Kashmir, while those of Urasa, Punach, and Rajapuri are spoken of in such a way as to show that they had kings of their own previous to their subjection by Kashmir.

Trigartta is repeatedly mentioned in the chronicles of Kashmir as an independent kingdom ; and its own history shows that one-half of the present petty states of the Jalandhar hills have sprung from the division of the possessions of a single family.

The following list gives the names and possessions of the various states attached to Kashmir, or the western division of the Alpine Panjab : —

Chiefs State
Khaka Bambas 1. Kashmir.
2. Gingal, on the Behat River
3. Muzafarabad, on the Behat River
4. Khagan, on the Kunihar River
Afghans 6. Garhi, on the Kunihar River
6. Rash, on Pakhli River
7. Dhantawar, on Dor River
8. Gandgarh, on Dor River
9. Darband, on the Indus River
10. Torbela, on the Indus River
Gakars 11. Pharwala, near Behat River
12. Sultanpur, on Behat River
l3. Khanpur, on Haro River

The Khaka-Bamba chiefs hold the valley of the Behat river below Barahmula, and the whole course

[p.132]: of the Kunihar river to the north-west of Kashmir. They are all Muhammadans, and are most probably the descendants of the early inhabitants of the country, who retired to their present position on the advance of the Afghan invaders.

The Afghan chiefs hold the valleys of the Pakhli and Dor rivers, to the south-west of Kashmir. They are all Muhammadans, and their settlement in this part of the country is of recent date. Abul Fazl mentions that before the time of Akbar, the raja of Pakhli was a tributary of Kashmir. He also states that Timur left a small body of troops in this district, whose descendants were still there in his time.1

The Gakar chiefs hold the lower valley of the Jhelam, and the upper course of the Haro river to the south-west of Kashmir. They are all Muhammadans ; but their conversion is comparatively recent, as their names were Indian down to the invasion of Timur. Their occupation of these districts is of very early date ; but they are Turanians, and not Arians, as none but a Gakar will intermarry with a Gakar, a practice that is utterly repugnant to Hinduism, which permits no man to marry one of his own tribe. The Gakars also occupy several portions of the eastern Doab, as Guliana, near Gujar Khan, and Bugial, under the lofty hill of Balnath. But these districts do not properly belong to the hills, although they were subject to Kashmir at the time of Hwen Thsang's visit in the seventh century.

The following list gives the names and positions of the various states attached to the central, or Jammu division of the Alpine Panjab : —

1 ' Ayin Akbari,' ii. 155.


Chiefs State
Hindus 1. Jammu, to east of Chenab River
2. Bhao, to east of Chenab River
Muhammadans 3. Rihasi, on Chenab River
4. Aknur, on Chenab River
5. Punach, on Punach River
6. Rajaori, on Tohi River
7. Kotali, on Punach River
8. Bhimbar, at foot of hills.
9. Khariali, near Bhimbar.
10. Kashtwar, on upper Chenab River
11. Bhadrwar, to south of Kastwar.
Hindus 12. Chaneni, to west of Bhadrwar.
13. Bandralta, to south of Chaneni.
14. Samba, to S.W. of Bandralta.
15. Jasrota, to south of Bandralta.
16. Tirikot, near Jasrota.
17. Mankot, to south of Bandralta.
18. Badwal, or Vaḍḍiwāsa.
19. Ballawar, or Bisohli.

The towns of Jammu and Bhao, which were founded by two brothers, are situated on opposite banks of the Tohi, a small stream that joins the Chenab at the foot of the hills. Jammu is mentioned several times in Muhammadan history, from the time when Timur forcibly converted the Raja down to the end of the last century. The three famous brothers of Ranjit Singh's court — Gulab Singh, Dhyan Singh, and Suchet Singh, belonged to a younger branch of this family, and the son of Gulab Singh now rules over Kashmir and the whole of the states in the western and central divisions of the Alpine Panjab.

The petty chiefs of Rihasi and Aknur were branches

[p.134]: of the Jammu family, on which they were generally dependent. Punach was sometimes independent ; but its proximity to Kashmir placed it at the mercy of its more powerful neighbour. Rajaori and Kotali were held in later times by two branches of the royal family of Kashmir, to which they were usually tributary. But in the middle ages, under the Hindu rulers, Kotali formed part of Punach, to which it naturally belonged as part of the same valley. Bhimbar and Khariali were divisions of the Chibh, or Chibhan, branch of the Somvansi Rajas of Kangra and Jalandhar. In early times the name of Bhimbar was little used, the common appellation being Chibhan, which is found in Sharifuddin's history of Timur, under the form of Jibhal. The conversion of the family to Muhammadanism is probably of late date, as Ferishta mentions Howns Raja of Bhimbar in a.h. 891, or A.D. 1486.1 But so many of these hill chiefs retained their Hindu names after they became Muhammadans, that the Hindu name alone cannot be taken as a decisive proof of his being unconverted. Kashtwar and Bhadrwar are situated on opposite banks of the upper Chenab river, to the south-east of Kashmir, to which they were generally subject. These nine chief-ships of the central division, added to the thirteen of the western division, form the twenty-two Muhammadan states which the popular belief assigns to the western half of the Alpine Panjab.

Of the eight remaining chiefships of this division I am not able to give much information, as many of them became extinct during the early period of Sikh rule, and all of them are now absorbed by the Jammu

1 Briggs, 'Ferishta,' iv. 483.

[p.135]: family in the great kingdom of Kashmir. Jasrota, in the outer range of hills, was once of some importance, and its chiefs intermarried with the other Rajput families of the Alpine Panjab; but I can find no mention of it in any of the histories. Ballawar and Badwal were certainly at one time under a single chief, as Kalasa, the son of Tukka, who is twice mentioned in the ' Raja Tarangini '1 as lord of Vallapura between 1028 and 1801, is found in the genealogical lists of both families. It is true that Vaddivasa is noticed in the same chronicle2 as a separate district at an earlier date, but as there is no mention of any chief, it may be inferred that it formed part of the small kingdom of Vallapura. As the names in the two genealogical lists differ from Kalasa downwards, it seems probable that the state may have been dismembered after his death. It is certain that he was mixed up with Kashmirian politics ; and as the contemporary Raja of the neighbouring state of Chamba was put to death by Ananta of Kashmir, I conclude that Ballawar must have been subjected at the same time.

I may remark that all the chiefs of the Central Division, whose genealogies I possess, trace their origin to the Surajvansi, or Solar Race, with the single exception of the intrusive Chibban of Bhimbar. The chiefs of Jammu, Jasrota, and Ballawar, with their offshoots, amounting together to eight of these petty states, all assert their descent from the Sun, a claim which is admitted by their Rajput neighbours.

The following list gives the names and positions of the various states attached to the eastern, or Jalandhar division of the Alpine Panjab.

1 ' Raja Tarangini,' vii. 220, 589, 2 Ibid., vi. 318, Nandigupta.


Vansha State
Somvansi 1. Kangra, or Katoch.
2. Guler, to S.W. of Kangra.
3. Jaswal, on Suhan River.
4. Datarpur, on lower Bias River.
5. Siba, on lower Bias River.
Surajvansi 6.Chamba, on Ravi River.
7. Kullu, on upper Bias River.
Pundir, or Pandayas. 8.Mandi, on middle Bias River.
9. Sukhet, to south of Mandi.
10. Nurpur, between Ravi and Bias River.
11. Kotila, to East of Nurpur.
12. Kotlehar.

Of these twelve states no less than five are mere subdivisions of the once rich kingdom of Jalandhar, which embraced the whole of the Doab, or plain country, between the Bias and Satlej, and all the hill country lying between the Ravi and the frontiers of Mandi and Sukket, to the south of the Dhaola-dhar mountains. This included Nurpur, Kotila, and Kotlehar ; and as Mandi and Sukhet were at first under one rule, there were originally only four chiefships in the eastern division of the Alpine Panjab, namely, Jalandhar, Chamba, Kullu, and Mandi.


Since the occupation of the plains by the Muhammadans, the ancient kingdom of Jalandhara has been confined almost entirely to its hill territories, which were generally known by the name of Kangra, after its most celebrated fortress. The district is also called Katoch, the meaning of which is unknown, and

[p.137]: Trigartta,1 which is the usual Sanskrit name found in the Puranas, and in the native chronicle of Kashmir.

In the seventh century Jalandhara is described by the Chinese pilgrim2 as about 1000 li, or 167 miles in length from east to west, and 800 li, or 133 miles in breadth from north to south. If these dimensions are even approximately correct, Jalandhar must then have included the state of Chamba on the north, with Mandi and Sukhet on the east, and Satadru on the south-east. As the last is the only district to the east of the Satlej, which is included in N. India, I infer that it must have belonged to the kingdom of Jalandhar. With the addition of these districts the size of the province will agree very well with the dimensions assigned to it by the Chinese pilgrim.

At the time of Hwen Thsang's visit, Jalandhar itself was the capital, which he describes as from 12 to 13 li, or upwards of 2 miles in circuit. Its antiquity is undoubted, as it is mentioned by Ptolemy as Kulindrine, or Ktulindrine, which should probably be corrected to Sulindrine, as the K and Σ are frequently interchanged in Greek manuscripts. According to the Padma Purana,3 the city of Jalandhara was the capital of the great Daitya king Jalandhara, who became so powerful by virtue of his austerities as to be invincible. At last, however, he was overcome by Siva, through a disgraceful fraud, and his body was devoured by the yoginis or female demons. But the conclusion of the legend is differently given in the

1 ' Hema-Koaha.' Jâlandharâs Trigarttâsyuh — " Jalandhara, that is Trigartta."

2 Julien's ' Hiouen Thsang,' ii. 202.

3 Uttara Khanda of the Padma Purana. Kennedy's ' Hindu Mythology,' p. 456.

[p.138]: local Purana,1 which states that he was overwhelmed and crushed to death by a mass of mountains which Siva placed upon him. Flames then sprang out of his mouth, which was under Jwala-mukhi ; his back was under the upper part of the Doab, which is still called Jalandhara-pitha, or Jalandhar-pith, by the people ; and his feet were under the lower part of the Doab at Multan. Akbar partially adopted this version of the legend when he named the different Doabs after the enclosing rivers, by calling the land between the Satlej and Bias the Doab-i-Bist Jalandhar or Bit Jalandhar, instead of the Sab Doab, which it should have been if he had placed the initial of the eastern river first, as he did in the names of the Bari and Chaj Doabs.

The royal family of Jalandhara and Kangra is one of the oldest in India, and their genealogy from the time of the founder, Susarma Chandra, appears to me to have a much stronger claim to our belief than any one of the long strings of names now shown by the more powerful families of Rajputana. All the different scions of this house claim to be of Somavansi descent ; and they assert that their ancestors held the district of Multan and fought in the Great War on the side of Duryodhan against the five Pandu brothers. After the war they lost their country, and retired under the leadership of Susarma Chandra to the Jalandhar Doab, where they established themselves, and built the stronghold of Kangra. The expedition of Alexander terminated on the banks of the Hyphasis, or Bias ; but he received the submission of Phegelas2 or

1 Jalaudhara Purana.

2 Diodorus, xvii. 51, " Phegaeus." Curtius, ix. 1, 3, " Phegelae erat gontis proximae rex."

[p.139]: Phegaeus, the king of the district, beyond the river, that is of the Jalandhar Doab. Towards the end of the fifth century, the kingdom of Trigartta was presented to Pravaresa by the Raja of Kashmir.1

In the seventh century, the Chinese pilgrim, Hwen Thsang, was courteously entertained for a whole month by Raja U-ti-to, or Udita,2 whom I would identify with Adima of the genealogical lists. One hundred and sixty years later, in an inscription dated A.D. 804, the Raja of Jalandhara is named Jaya Chandra, who is the Jaya Malla Chandra of the lists, the seventh in descent from Adima. Lastly, Avanta, king of Kashmir, from A.D. 1028 to 1081, married two daughters of Indu Chandra, 3 Raja of Jalandhara, who is the Indra Chandra of the genealogical lists of Kangra. These instances are sufficient to show that Jalandhara existed as an independent State for many centuries before the Muhammadan conquest.

The smaller chiefships of Guler, Jaswal, Datarpur, and Siba, are offshoots from the parent stem of Kangra. The independence of Guler, or Haripur, was established by Hari Chandra, about A.D. 1400, when he yielded Kangra to his younger brother, Karmma Chandra. The date of the foundation of the other principalities is unknown, but I believe that they were always tributary to the parent state until the time of the Muhammadans, when the capture of Kangra by Mahmud of Ghazni afforded them an opportunity of asserting their independence.

The French traveller Thevenot,4 in his account of the dominions of the Emperor of Delhi, mentions

1 'Raja Tarangini,' iii. 100. 2 Julien's ' Hiouen Thsang,' i. 261. 3 ' Raja Tarangini,' vii. 150. 4 ' Travels,' part iii. c. 37.

[p.140]: that " there are many Rajas who own not the authority of the Great Mogul." But the territories of these Rajas must have been far in the interior of the hills, as we know that the chiefs of all the outer hills were subjected by the Mogul emperors. Thevenot specially mentions the province of " Ayoud, or Haoud" as containing " the most northern countries that belong to the Great Mogul, as Caucares, Bankish, Nagarcut, Siba, and others."

The Caucares must be the Gakars who hold the lower hills to the west of the Jhelam. Terry1 calls them Kakares, and their principal cities Dekalee and Furhola (or Dangali and Pharwala). The Bankish are the Banchish of Terry,2 I whose " chief city, called Bishur (Peshawar) lyeth east (read west) somewhat southerly from Chishmere, from which it is divided by the river Indus." Nagarcut is Kangra or Nagarkot, which is mentioned under the same name by Abu Rihan,3 who was present at its capture by Mahmud of Ghazni. Siba is not as we might suppose, the small state in the neighbourhood of Kangra, but a district on the Ganges, of which the chief city, according to Terry, was " Hardware (or Haridwara), where the river Ganges, passing through or amongst large rocks, makes presently after a pretty full current." From these accounts it is clear that the whole of the states in the lower hills, from Peshawar on the west to the Ganges on the east, were subject to the emperor of Delhi. Regarding the general name of Ayoud, or Haoud, which Thevenot applies to them, I can only conjecture that it may be some corrupt form of Himavat, or Himwat, --- 1 ' Voyage to East India,' p. 88. 2 Ibid., p. 81 : London, 1655. 3 ' Fragments Arabes, etc.,' 149.

[p.141]: one of the well-known names of the Hmalaya mountains, which the Greeks have preserved under the two different forms of Emodos and Imaus.

Champa or Chamba

Chamba is a large district, which includes the valleys of all the sources of the Ravi, and a portion of the upper valley of the Chenab, between Lahul and Kashtwar. It is not mentioned by Hwen Thsang, and therefore, probably included by him within the limits of Kashmir The ancient capital was Varmmapuri oi Barmawar, on the Budhil river, where many fine temples, and a brazen bull, of life size, still exist to attest the wealth and piety of its early rulers. According to the inscriptions these works belong to the ninth and tenth centuries. The country is frequently mentioned in the native chronicle of Kashmir, under the name of Champa, and each notice is confirmed by the local genealogies. Between A.D. 1028 and 1031 the district was invaded by Ananta of Kashmir,1 when the native Raja, named Sala, was defeated and put to death. His son founded a new capital, Champapura, called after the goddess Champavati Devi, which, under the name of Chamba, is still the chief place in the district. The Rajas of Kashmir after-wards intermarried with the Chamba family ;2 and during the troubles that followed the Muhammadan invasions this petty state became independent, and remained so until reduced by Gulab Singh, early in the present century.

1 Briggs's 'Ferishta,' i. 283. The Gakars inhabited the banks of the Nilab (or Indus) up to the foot of the mountains of Siwalik.

2 'Raja Tarangini," vii. 218. Ibid., vii. 589, 1520; viii. 1092.


[p.142]: The kingdom of Kiu-lu-to is placed by Hwen Thsang at 700 li, or 117 miles, to the north-east of Jalandhar,1 which corresponds exactly with the position of the district of Kullu, in the upper valley of the Byas river. The Vishnu Purana2 mentions a people called Uluta, or Kuluta, who are most probably the same as the Kaulutas of the ' Ramayana' and the ' Brihat Sanhita.3 As this form of the word agrees precisely with the Chinese Kiuluto, I conclude that the modern Kullu, must be only an abbreviation of the ancient name. The district is stated to be 3000 li, or 500 miles, in circuit, and entirely surrounded by mountains. The size is very much exaggerated for the present restricted limits of Kullu ; but as the ancient kingdom is said by the people themselves to have included Mandi and Sukhet on the west, and a large tract of territory to the south of the Satlej, it is probable that the frontier measurement of 500 miles may be very near the truth if taken in road distance.

The present capital of the valley is Sultanpur ; but the old capital of Makarsa is still called Nagar, or the city, by which name it is most generally known. Hwen Thsang states that gold, silver, and copper are all found in the district, which is only partially true, as the amount of gold to be obtained by washing is very small, and the silver and copper mines have long been abandoned.

To the north-east of Kullu Hwen Thsang places the district of Lo-hu-lo, which is clearly the Lho-yal of

1 Julien's 'Hiouen Thsang,' ii. 203.

2 Wilson's 'Vishnu Purana,' edited by Hall, ii. 3, vol. ii. p. 174.

3 Kern's ' Brihat Sanhita,' xiv. 29.

[p.143]: the Tibetans, and the Lahul of the people of Kullu and other neighbonring states. Still further to the north he places the district of Mo-lo-so, which, from his position, must certainly be Ladak. I would, therefore, alter the Chinese name to Mo-lo-po, which is an exact transcript of Mar-po, the actual name of the province of Ladak, as Mar-po-yul, or the "Red district," in allusion to the general appearance of its soil and mountains. The Chinese syllables so and po are so much alike that they are frequently interchanged, as in the well-known name of Salatura, the birth-place of Panini, which is given in the original Chinese of Hwen Thsang's travels as Po-lo-lu-lo, or Palatura.

Mandi and Sukhet

The petty chiefships of Mandi and Sukhet were originally a single state, bounded by Kangra on the west and Kullu on the east, and by the Dhaoladhar mountains on the north and the Satlej on the south. Mandi means the "market;" and its favourable position on the Bias river, at the junction of the two roads from the west and south, must have ensured its early occupation, which was rendered prosperous and lasting by the existence of valuable mines of iron and black salt in its immediate vicinity.

Nurpur, or Pathaniya.

The town of Nurpur derives its name from the celebrated Nur Jahan, the wife of the emperor Jahanjir. Its original name was Dahmari, or Dahmala ; or as Abul Fazl writes, Dahmahri, although he mentions no fort. The people pronounce the name as if written Dahmeri. In the ' Tarikh-i-Alfi. ' it is called Damal, an dis described as "situated on the summit of a high hill,

[p.144]: on the borders of Hindustan." The fort was taken after a long siege by Ibrahim Ghaznavi. The name of the district is Pathawat, and the old capital in the plains was called Pathian, or Pathiankot, which is now slightly altered to Pathankot. But the name is derived from the Pathan tribe of Hindu Rajputs, and not from the well-known Muhammadan Pathans, or Afghans. The Raja was imprisoned in 1815 by Ranjit Singh, who took possession of his country.

The petty chief of Kotila, to the east of Nurpur, who was a scion of the Pathaniya family, was seized about the same time, and his estate incorporated with the Sikh dominions.

Kotlehar is a petty state in the Jaswal Dun, to the south-east of Jwala-Mukhi. It was generally a dependency of Kangra.


The district of She-to-tu-lo or Satadru, is described by the Chinese pilgrim1 as 2000 li, or 333 miles in circuit, with a large river forming its western boundary. The capital is placed at 700 li, or 117 miles, to the south of Kullu, and 800 li, or 133 miles, to the north-east of Bairat. But there is a mistake in one of these numbers, as the distance between the capital of Kullu and Bairat is 336 miles, measured direct on the map, or not less than 360 miles, by road. There is a deficiency, therefore, in one of the distances of about 110 miles, or nearly 700 li, in a direct line between the two places, or of about 150 miles, or nearly 1000 li, in the detour, as shown by his bearings. Now it is remarkable that there is a deficiency

1 Julien's ' Hiouen Thsang,' ii. 205.

[p.145]: of about the same amount in the return journey along a parallel line of road, from Mathura to Thanesar, which the pilgrim makes only 500 li, or 83 miles,1 instead of 1200 li, or 200 miles, the actual distance being 199 miles. As it would seem that both routes, for some unknown reason, had been subjected to the same amount of curtailment, it is probable that the deficiency in the western line will lie in the southern portion between Satadru and Bairat, which is contiguous to the parallel line between Mathura and Thanesar. I would, therefore, increase the distance between the two former places by 150 miles, or in round numbers 1000 li, which would make the total distance 283 miles, or nearly 1800 li, instead of 800 li. Taking this corrected distance from Bairat, and the recorded distance of 117 miles south from Kullu, the position of Satadru will correspond almost exactly with the large city of Sarhind, which both history and tradition affirm to be the oldest place in this part of the country.

The present ruins of Sarhind consist almost entirely of Muhammadan buildings of a late period ; but it must have been a place of some consequence in the time of the Hindus, as it was besieged and captured by Muhammad Ghori, the first Mussulman king of Delhi. 2 The name of Sarhind, or " frontier of Hind" is popularly said to have been given to the city at an earlier period, when it was the boundary town between the Hindus and the later Muhammadan kingdom of Ghazni and Lahor. But the name is probably

1 Julien's ' Hiouen Thsang,' i. 104, and ii. 211. 2 Dowson's edition of Sir H. Elliot's ' Muhammadan Historians of India,' ii. 295.

[p.146]: much older, as the astronomer Varaha Mihira mentions the Sairindhas1 immediately after the Kulutas, or people of Kullu, and just before Brahmapura, which, as we learn from the Chinese pilgrim, was the capital of the hill country to the north of Haridwar. The Sairindhas, or people of Sirindha, must, therefore, have occupied the very tract of country in which the present Sarhind is situated, and there can be little doubt that the two names are the same. But the geographical list of Varaha Mihira is copied almost verbatim from that of the still earlier astronomer Parasara, who is believed to have flourished not later than the first century after Christ.2

If we apply the correction of 110 miles, or about 700 li, to the northern half of the line between Kullu and Satadru, the position of the latter will be brought down to Hansi, which is an ancient fortified city of even greater strength and reputation than Sarhind. But as Hwen Thsang specially notes that the territory of Satadru was only 2000 li, or 333 miles, in circuit, and that it was bounded on the west by a great river, which can only be the Satlej or Satadru, it is quite impossible that Hansi could be the place intended, as it is upwards of 130 miles distant from the nearest point of that river.

The position of the celebrated fortress of Bhatner would suit the description of a small district bounded on the west by the Satlej, and would also agree with the corrected distance from Kullu : but the direction is south-west instead of south, and the distance from Bairat is upwards of 200 miles, instead of 800 li, or

1 Dr. Kern's edition of the ' Brihat Sanhita,' b. xiv. 29, 30.

2 Kern's Preface to the ' Brihat Sanhita,' p. 32.

[p.147]: 133 miles, as stated by the pilgrim. The bearing of Bairat is, however, in favour of Bhatner, as the pilgrim's south-west is certainly a mistake for south-east, otherwise the distance of Bairat from Mathura would be nearly 1600 li, or 250 miles, instead of 500 li, or 83 miles, as recorded. If we might read 1500 li instead of 500 li, the relative positions of Bhatner and Bairat would correspond very well with the pilgrim's account, as the road distance between the two places, via Hansi, is about 250 miles. It is quite possible also that there may be a mistake in the initial Chinese character. She or ṣa, which is very much like Po or Bha ; and if so, the Chinese syllables would represent Bhatasthala, or Bhatner. The latter name means the " fortress of the Bhatis," but the town itself was called Band, or Bando, which was probably the contracted form of Bhatasthala, just as Maru is now the common contracted form of Marusthala. But in spite of these plausible agreements both in name and in position, I am inclined to think that Sarhind must be the place indicated by the pilgrim as the capital of the ancient district of Satadru, This conclusion is strengthened by the pilgrim's statement that the country produced gold, which, so far as I know, can only apply to the lower hills lying to the north of Sarhind, where gold is still found in some of the smaller affluents of the Satlej.

Accepting Sarhind as the capital of Satadru, the boundaries of the district may be determined approximately from its size. On the west and north it was bounded by the Satlej for upwards of 100 miles from the neighbourhood of Simla to Tihara, below Ludiana. On the south the boundary extended for about 100

[p.148]: miles from Tihara to Ambala, and on the east for about the same distance, from Ambala to Simla. The circuit thus described embraces a considerable portion of the hill states to the west and south of Simla, together with the districts of Sarhind proper and Ludiana in the plains. As it is the only district lying to east of the Satlej that is included within the limits of Northern India, I infer that it must have been a dependency of the neighbouring state of Jalandhar.

Back to Index of the Book