Transcriber’s Note:

The text is annotated with numerous footnotes, which were numbered sequentially on each page. On occasion, a footnote itself is annotated by a note. In the previous two volumes, these were referenced using symbols (e.g. *, †), which have been converted to an alphabetic sequence (e.g., A, B). In this volume, however, the ‘subnotes’ appear in one instance as numerals. For the sake of consistency, the convention used in the previous volumes has been adopted.

Since there are over 900 notes in this volume, they have been gathered at each chapter’s end, and resequenced for each chapter. Links are provided to navigate from the reference to the note, and back.

The notes are a combination of those of the author, and of the editor of this edition. The text of the latter are enclosed in square brackets. The bold-faced phrases that begin each topic were also added by the editor, and spelling of Hindi or Sanskrit words may vary between those phrases and the author’s text.

The pagination of the original edition, published in the 1820's, is preserved for ease of reference by including those page numbers in the text, also enclosed in square brackets.

Crooke’s plan for the renovation of the Tod’s original text, including a discussion of the transliteration of words other than English, is given in detail in the [Preface].

Several tables spanned multiple pages, with sums totaled before the page break as “Carried forward”, and repeated on the following page. These have been removed, given the nature of the current text.

Minor errors, attributable to the printer, have been corrected. Given the history of the text, it was thought best to leave all orthography as printed.

Please see the transcriber’s [note] at the end of this text for details regarding the handling of any textual issues encountered during its preparation.

A complex genealogical chart appears on p. [1457], inserted in mid-paragraph spanning pp. 1456 and 1458. It has been moved to precede that paragraph. The chart was not conducive to rendering in this format. It has been approximated in the text, and a link provided to an image of the original page.

The cover image has been amended to include identifying information, and is added, as amended, to the public domain.

Any corrections are indicated using an underline highlight. Placing the cursor over the correction will produce the original text in a small popup.

Any corrections are indicated as hyperlinks, which will navigate the reader to the corresponding entry in the corrections table in the note at the end of the text.

ANNALS AND ANTIQUITIES

OF RAJASTHAN

COLONEL TOD AND HIS JAIN GURU.
(From a painting said to be the work of the Author’s native artist, Ghāsi.)
Frontispiece.

ANNALS AND ANTIQUITIES
OF
RAJASTHAN

OR THE CENTRAL AND WESTERN

RAJPUT STATES OF INDIA

BY

Lieut.-Col. JAMES TOD

LATE POLITICAL AGENT TO THE WESTERN RAJPUT STATES

EDITED WITH AN INTRODUCTION AND NOTES BY

WILLIAM CROOKE, C.I.E.

HON. D.SC. OXON., B.A., F.R.A.I.

LATE OF THE INDIAN CIVIL SERVICE

IN THREE VOLUMES

VOL. III.

HUMPHREY MILFORD

OXFORD UNIVERSITY PRESS

LONDON EDINBURGH GLASGOW NEW YORK

TORONTO MELBOURNE BOMBAY

1920

CONTENTS

ILLUSTRATIONS

ANNALS AND ANTIQUITIES

OF RAJASTHAN

BOOK VIII
SKETCH OF THE INDIAN DESERT

CHAPTER 1

Having never penetrated personally farther into the heart of the desert than Mandor, the ancient capital of all Marusthali, the old castle of Hissar on its north-eastern frontier, and Abu, Nahrwala, and Bhuj, to the south, it may be necessary, before entering upon the details, to deprecate the charge of presumption or incompetency, by requesting the reader to bear in mind that my parties of discovery have traversed it in every direction, adding to their journals of routes living testimonies of their accuracy, and bringing to me natives of every thal from Bhatner to Umarkot, and from Abu to Aror.[^[1]] I wish it, however, to be clearly understood, that I look upon this as a mere outline, which, by showing what might be done, may stimulate further research; but in the existing dearth of information on the subject I have not hesitated to send it forth, with its almost inevitable errors, as (I trust) a pioneer to more extended and accurate knowledge.

After premising thus much, let us commence with details, which, but for the reasons already stated, should have been comprised in the geographical portion of the work, and which, though irrelevant to the historical part, are too important to be [290] thrown into notes. I may add, that the conclusions formed, partly from personal observation, but chiefly from the resources described above, have been confirmed by the picture drawn by Mr. Elphinstone of his passage through the northern desert in the embassy to Kabul, which renders perfectly satisfactory to me the views I before entertained. It may be well, at this stage, to mention that some slight repetitions must occur as we proceed, having incidentally noticed many of the characteristic features of the desert in the Annals of Bikaner, which was unavoidable from the position of that State.[State.]

Description of the Desert.

Boundaries of the Desert.

We may here repeat the tradition illustrating the geography of the desert, i.e. that in remote ages it was ruled by princes of the Panwar (Pramara) race, which the sloka, or verse of the bard, recording the names of the nine fortresses (Nau-koti Maru-ki), so admirably adapted by their position to maintain these regions in subjection, further corroborates. We shall divest it of its metrical form, and begin with Pugal, to the north; Mandor, in the centre of all Maru; Abu, Kheralu, and Parkar, to the south; Chhotan, Umarkot, Aror, and Lodorva, to the west; the possession of which assuredly marks the sovereignty of the desert. The antiquity of this legend is supported by the omission of all modern cities, the present capital of the Bhattis not being mentioned. Even Lodorva and Aror, cities for ages in ruins, are names known only to a few who frequent the desert; and Chhotan and Kheralu, but for the traditional stanzas which excited our research, might never have appeared on the map.

Natural Divisions of the Desert.

The whole of Bikaner, and that part of Shaikhavati north of the Aravalli, are comprehended in the desert. If the reader will refer to the map, and look for the town of Kanod,[^[2]] within the British frontier, he will see what Mr. Elphinstone considered as the commencement of the desert, in his interesting expedition to Kabul.[^[3]] “From Delly to Canound (the Kanorh of my map), a distance of one hundred miles is through the British dominions, and need not be described. It is sufficient to say that the country is sandy, though not ill cultivated. On approaching Canound, we had the first specimen of the desert, to which we were looking forward with anxious curiosity. Three miles before reaching that place we came to sand-hills, which at first were [292] covered with bushes, but afterwards were naked piles of loose sand, rising one after another like the waves of the sea, and marked on the surface by the wind like drifted snow. There were roads through them, made solid by the treading of animals; but off the road our horses sunk into the sand above the knee.” Such was the opening scene; the route of the embassy was by Singhana, Jhunjhunu, to Chum, when they entered Bikaner. Of Shaikhavati, which he had just left, Mr. Elphinstone says: “It seems to lose its title to be included in the desert, when compared with the two hundred and eighty miles between its western frontier and Bahawulpoor, and, even of this, only the last hundred miles is absolutely destitute of inhabitants, water, or vegetation. Our journey from Shekhavati to Poogul was over hills and valleys of loose and heavy sand. The hills were exactly like those which are sometimes formed by the wind on the seashore, but far exceeding them in height, which was from twenty to a hundred feet. They are said to shift their position and alter their shapes according as they are affected by the wind; and in summer the passage is rendered dangerous by the clouds of moving sand; but when I saw the hills (in winter), they seemed to have a great degree of permanence, for they bore grass, besides phoke, the babool, and bair or jujube, which altogether give them an appearance that sometimes amounted to verdure. Amongst the most dismal hills of sand one occasionally meets with a village, if such a name can be given to a few round huts of straw, with low walls and conical roofs, like little stacks of corn.” This description of the northern portion of the desert, by an author whose great characteristics are accuracy and simplicity, will enable the reader to form a more correct notion of what follows.[^[4]]

With these remarks, and bearing in mind what has already been said of the physiography of these regions, we proceed to particularize the various thals and oases in this ‘region of death.’ It will be convenient to disregard the ancient Hindu geographical division, which makes Mandor the capital of Marusthali, a distinction both from its character and position better suited to Jaisalmer, being nearly in the centre of what may be termed entire desert. It is in fact an oasis, everywhere insulated by immense masses of thal, some of which are forty miles in breadth, without the trace of man, or aught that could subsist him. From Jaisalmer we shall pass to Marwar, and without crossing the Luni, describe Jalor and Siwanchi; then conduct the [293] reader into the almost unknown Raj of Parkar and Virawah,[^[5]] governed by princes of the Chauhan race, with the title of Rana. Thence, skirting the political limits of modern Rajputana, to the regions of Dhat and Umra-sumra, now within the dominion of Sind, we shall conclude with a very slight sketch of Daudputra, and the valley of the Indus. These details will receive further illustration from the remarks made on every town or hamlet diverging from the ‘hill of Jaisal’ (Jaisalmer). Could the beholder, looking westward from this ‘triple-peaked hill,’[^[6]] across this sandy ocean to the blue waters (Nilab)[^[7]] of the Indus, embrace in his vision its whole course from Haidarabad to Uchh, he would perceive, amidst these valleys of sand-hills, little colonies of animated beings, congregated on every spot which water renders habitable. Throughout this tract, from four hundred to five hundred miles in longitudinal extent, and from one hundred to two hundred of diagonal breadth, are little hamlets, consisting of the scattered huts of the shepherds of the desert, occupied in pasturing their flocks or cultivating these little oases for food. He may discern a long line of camels (called kitar, a name better known than either kafila or karwan), anxiously toiling through the often doubtful path, and the Charan conductor, at each stage, tying a knot on the end of his turban. He may discover, lying in ambush, a band of Sahariyas, the Bedouins of our desert (sahra),[^[8]] either mounted on camels or horses, on the watch to despoil the caravan, or engaged in the less hazardous occupation of driving off the flocks of the Rajar or Mangalia shepherds, peacefully tending them about the tars or bawas, or hunting for the produce stored amidst the huts of the ever-green jhal,[^[9]] which serve at once as grain-pits and shelter from the sun. A migratory band may be seen flitting with their flocks from ground which they have exhausted, in search of fresh pastures:

And if the following day they chance to find

A new repast, or an untasted spring,

Will bless their stars, and think it luxury!

Or they may be seen preparing the rabri, a mess quite analogous to the kouskous of their Numidian brethren, or quenching their thirst from the Wah of their little oasis, of which they maintain sovereign possession so long as the pasture lasts, or till they come in conflict with some more powerful community.

Oasis.

At the risk of somewhat of repetition, we must here point out the few grand features which diversify this sea of sand, and after defining the difference between rui and thal, which will frequently occur in the itinerary, at once plunge in medias res.

The Lost River of the Desert.

The Lūni River.

The Rann of Cutch.

Thal, Rūi.

Thal of the Luni.

Jālor.

Sonagir, the ‘golden mount,’ is the more ancient name of this castle, and was adopted by the Chauhans as distinctive of their tribe, when the older term, Mallani, was dropped for Sonigira. Here they enshrined their tutelary divinity, Mallinath, ‘god of the Malli,’ who maintained his position until the sons of Siahji entered these regions, when the name of Sonagir was exchanged for that of Jalor, contracted from Jalandharnath, whose shrine is about a coss west of the castle. Whether Jalandharnath [297], the ‘divinity of Jalandhar,’ was imported from the Ganges, or left as well as the god of the Malli by the ci-devant Mallanis, is uncertain: but should this prove to be a remnant of the foes of Alexander, driven by him from Multan,[^[19]] its probability is increased by the caves of Jalandhar (so celebrated as a Hindu pilgrimage even in Babur’s time) being in their vicinity. Be this as it may, the Rathors, like the Roman conquerors, have added these indigenous divinities to their own pantheon. The descendants of the expatriated Sonigiras now occupy the lands of Chitalwana, near the furca of the Luni.

Jalor comprehends the inferior districts of Siwanchi, Bhinmal, Sanchor, Morsin, all attached to the khalisa or fisc; besides the great pattayats, or chieftainships, of Bhadrajan, Mewa, Jasola, and Sindari—a tract of ninety miles in length, and nearly the same in breadth, with fair soil, water near the surface, and requiring only good government to make it as productive as any of its magnitude in these regions, and sufficient to defray the whole personal expenses of the Rajas of Jodhpur, or about nine lakhs of rupees; but in consequence of the anarchy of the capital, the corruption of the managers, and the raids of the Sahariyas of the desert and the Minas of Abu and the Aravalli, it is deplorably deteriorated. There are several ridges (on one of which is the castle) traversing the district, but none uniting with the table-land of Mewar, though with breaks it may be traced to near Abu. In one point it shows its affinity to the desert, i.e. in its vegetable productions, for it has no other timber than the jhal, the babul, the karil, and other shrubs of the thal.

The important fortress of Jalor, guarding the southern frontier of Marwar, stands on the extremity of the range extending north to Siwana. It is from three to four hundred feet in height, fortified with a wall and bastions, on some of which cannon are mounted. It has four gates; that from the town is called the Suraj-pol, and to the north-west is the Bal-pol (‘the gate of Bal,’ the sun-god), where there is a shrine of the Jain pontiff, Parsvanath. There are many wells, and two considerable baoris, or reservoirs of good water, and to the north a small lake formed by damming up the streams from the hills; but the water seldom lasts above half the year. The town [298], which contains three thousand and seventeen houses, extends on the north and eastern side of the fort, having the Sukri flowing about a mile east of it. It has a circumvallation as well as the castle, having guns for its defence; and is inhabited by every variety of tribe, though, strange to say, there are only five families of Rajputs in its motley population. The following census was made by one of my parties, in A.D. 1813:

The general accuracy of this census was confirmed.

Sīwāna.

Machola, Morsin.

Bhīnmāl, Sānchor.

Bhadrājan.

Mewa.

Bālotra, Tīlwāra.

Īndhāvati.

Gūgadeo ka Thal.

Tararoi.

Khawar.

Mallināth, Bārmer.

Kherdhar.

Jūna Chhotan.

Nagar Gurha.

Here terminate our remarks on the thals of western Marwar, which, sterile as it is by the hand of Nature, had its miseries completed by the famine that raged generally throughout these regions in S. 1868 (A.D. 1812), and of which this[^[32]] is the third year. The disorders which we have depicted as prevailing at the seat of government for the last thirty years, have left these remote regions entirely to the mercy of the desert tribes [304], or their own scarce less lawless lords: in fact, it only excites our astonishment how man can vegetate in such a land, which has nothing but a few sars, or salt-lakes, to yield any profit to the proprietors, and the excellent camel pastures, more especially in the southern tracts, which produce the best breed in the desert.

[1]. The journals of all these routes, with others of Central and Western India, form eleven moderate-sized folio volumes, from which an itinerary of these regions might be constructed. It was my intention to have drawn up a more perfect and detailed map from these, but my health forbids the attempt. They are now deposited in the archives of the Company, and may serve, if judiciously used, to fill up the only void in the great map of India, executed by their commands.

[2]. [Kānod Mohindargarh in Patiāla State (IGI, xvii. 385).]

[3]. It left Delhi October 13, 1808.

[4]. “Our marches,” says Mr. Elphinstone, “were seldom very long. The longest was twenty-six miles, and the shortest fifteen; but the fatigue which our people suffered bore no proportion to the distance. Our line, when in the closest order, was two miles long. The path by which we travelled wound much, to avoid the sand-hills. It was too narrow to allow of two camels going abreast; and if an animal stepped to one side, it sunk in the sand as in snow,” etc. etc.—Account of the Kingdom of Caubul, ed. 1842, vol. i. p. 11.

[5]. [In Sind, on the N. shore of the Great Rann, about 10 miles from Nagar-Pārkar.]

[6]. Trikuta, the epithet bestowed on the rock on which the castle of Jaisalmer is erected.

[7]. A name often given by Ferishta to the Indus.

[8]. [As has been already stated, Sahariya has no connexion with Arabic Sahra, ‘desert.’]

[9]. [Jhāl, of which there are two varieties, large and small, Salvadora persica and S. oleoides.]

[10]. When I penned this conjectural etymology, I was not aware that any speculation had been made upon this word: I find, however, the late M. Langlés suggested the derivation of oasis (variously written by the Greeks αὔασις, ἴασις and υἅσις, ὄασις, [αὔασις is the only other recognized form]) from the Arabic واح: and Dr. Wait, in a series of interesting etymologies (see Asiatic Journal, May 1830), suggests वसि, vasi from वस, vas, ‘to inhabit.’ Vasi and ὕασις quasi vasis are almost identical. My friend, Sir W. Ouseley, gave me nearly the same signification of وادي, Wadi, as appears in Johnson’s edition of Richardson, namely, a valley, a desert, a channel of a river—a river; وادي, wadi-al-kabir, ‘the great river,’ corrupted into Guadalquiver, which example is also given in d’Herbelot (see Vadi Gehennem), and by Thompson, who traces the word water through all the languages of Europe—the Saxon waeter, the Greek ὔδωρ, the Islandic udr, the Slavonic wod (whence woder and oder, ‘a river’): all appear derivable from the Arabic wad, ‘a river’—or the Sanskrit wah; and if Dr. W. will refer to p. [1322] of the Itinerary, he will find a singular confirmation of his etymology in the word bas (classically vas) applied to one of these habitable spots. The word basti, also of frequent occurrence therein, is from basna, to inhabit; vasi, an inhabitant; or vas, a habitation, perhaps derivable from wah, indispensable to an oasis! [The New English Dict. gives Lat. oasis, Greek ὄασις, apparently of Egyptian origin; cf. Coptic ouahe (whence Egyptian Arabic wāh), ‘dwelling-place, oasis,’ from ouih, ‘to dwell.’]

[11]. [See IGI, xii. 212 f.; E. H. Aitken, Gazetteer of Sind, 4; Calcutta Review, 1874; JRAS, xxv. 49 ff.]

[12]. [The derivation of Pārkar is unknown; that suggested in the text is impossible.]

[13]. [Nārāyansar, an important place of pilgrimage, with interesting temples, is situated at the Kori entrance of the W. Rann (BG, v. 245 ff.).]

[14]. [Or irina, Yule, Hobson-Jobson, 2nd ed. 774.]

[15]. [Equus hemionus (Blanford, Mammalia of India, 470 f.; Job xxxix. 5 ff.).]

[16]. “The greatest breadth of the valley of the Nile is four leagues, the least, one”; so that the narrowest portion of the valley of Sind equals the largest of the Nile. Egypt alone is said to have had eight millions of inhabitants; what then might Sind maintain! The condition of the peasantry, as described by Bourrienne, is exactly that of Rajputana; “The villages are fiefs belonging to any one on whom the prince may bestow them; the peasantry pay a tax to their superior, and are the actual proprietors of the soil; amidst all the revolutions and commotions, their privileges are not infringed.” This right (still obtaining), taken away by Joseph, was restored by Sesostris.

[17]. Another salt river.

[18]. [The Chauhān Rāo Kīrttipāl took it from the Pramāras towards the end of the twelfth century, and Kānardeo Chauhān lost it to Alāu-d-dīn (Erskine iii. A. 199 f.). In Briggs’ translation of Ferishta (i. 370) the place is called Jalwar, and the King Nāhardeo.]

[19]. Multan and Juna (Chhotan, qu. Chauhan-tan?) have the same signification, ‘the ancient abode,’ and both were occupied by the tribe of Malli or Mallani, said to be of Chauhan race; and it is curious to find at Jalor (classically Jalandhar) the same divinities as in their haunts in the Panjab, namely, Mallinath, Jalandharnath, and Balnath. Abu-l-Fazl[Abu-l-Fazl] says, “The cell of Balnath is in the middle of Sindsagar”; and Babur (Elliot-Dowson ii. 450, iv. 240, 415, v. 114, Āīn, ii. 315) places “Balnath-jogi below the hill of Jud, five marches east of the Indus,” the very spot claimed by the Yadus, when led out of India by their deified leader Baldeo, or Balnath.

[20]. [Bhojak, ‘a feeder,’ a term usually applied to those Brāhmans who are fed after a death, in order to pass on the food to the spirit.]

[21]. [Ferishta (i. 369) calls the Rāja Sītaldeo; Amīr Khusru (Elliot-Dowson iii. 78, 550, v. 166) Sutaldeo.]

[22]. [The population of these towns is now respectively 4545 and 2066.]

[23]. [The old name was Srīmāl or Bhillamāla, which Erskine (iii. A. 194) identifies with Pi-lo-mo-lo of Hiuen Tsiang. But Beal (Buddhist Records of the Western World, ii. 270) transliterates this name as Bālmer or Bārmer.]

[24]. [For the Sāchora or Sānchora Brāhmans see BG, ix. Part i. 18; Erskine iii. A. 84.]

[25]. [Tīlwāra is about 10 miles W. of Bālotra.]

[26]. It is asserted by the natives to be caused by a small thread-like worm, which also forms in the eyes of horses. I have seen it in the horse, moving about with great velocity. They puncture and discharge it with the aqueous humour.

[27]. [The name Tararoi seems to have disappeared from the maps, the tract being now known as Sānkra.]

[28]. [Rāmdeora is 12 miles N. of Pokaran. The saint is commonly called Rāmdeoji or Rāmsāh Pīr.]

[29]. [Bārmer, the ancient name of which is said to be Bāhadamer, ‘hill fort of Bāhada,’ is 130 miles W. of Jodhpur city; its present population is 6064. Mallināth was son of Rāo Salkha, eighth in descent from Siāhji, founder of Mārwār State.]

[30]. Named in all probability, from the superabundant tree of the desert termed Khair, and dhar, ‘land.’ It is also called Kheralu, but more properly Kherala, ‘the abode of Khair’; a shrub of great utility in these regions. Its astringent pods, similar in appearance to those of the laburnum, they convert into food. Its gum is collected as an article of trade; the camels browse upon its twigs, and the wood makes their huts. [Kher is a ruined village, not far from Jasol, at the point where the Lūni River turns eastward. Kherālu has disappeared from modern maps, if it be not a mistake for Kerādu, where there are interesting temples (ASR, West Circle, March 31, 1907, pp. 40-43; Erskine iii. A. 201).]

[31]. [Khair, Acacia catechu; Khejra, Prosopis spicigera; Karīl, Capparis aphylla; Khep, Crotolaria burhia; Phog, Calligonum polygonoides.]

[32]. That is, 1814. I am transcribing from my journals of that day, just after the return of one of my parties of discovery from these regions, bringing with them natives of Dhat, who, to use their own simple but expressive phraseology, “had the measure of the desert in the palm of their hands”; for they had been employed as kasids, or messengers, for thirty years of their lives. Two of them afterwards returned and brought away their families, and remained upwards of five years in my service, and were faithful, able, and honest in the duties I assigned them, as jamadars of daks, or superintendents of posts, which were for many years under my charge when at Sindhia’s court, extending at one time from the Ganges to Bombay, through the most savage and little-known regions in India. But with such men as I drilled to aid in these discoveries, I found nothing insurmountable. [The famine of 1812-13 was the most calamitous of the earlier visitations (Erskine iii. A. 125).]

CHAPTER 2

The Chauhān Rāj.

History of the Chauhāns.

Although nominally a single principality, the chieftain of Parkar pays little, if any, submission to his superior of Virawah. Both of them have the ancient Hindu title of Rana, and are said at least to possess the quality of hereditary valour, which is synonymous with Chauhan. It is unnecessary to particularize the extent in square miles of thal in this raj, or to attempt to number its population, which is so fluctuating; but we shall subjoin a brief account of the chief towns, which will aid in estimating the population of Marusthali. We begin with the first division.

Chief Towns.

Nagar Pārkar.

Bakhasar.

Tharād.

Face of the Chauhān Rāj.

Water Production.

Inhabitants.

Kolis and Bhils.

Pital is the chief husbandman of this region, and, with the Bania, the only respectable class. They possess flocks, and are also cultivators, and are said to be almost as numerous as either the Bhils or Kolis. The Pital is reputed synonymous with the Kurmi of Hindustan and the Kulambi of Malwa and the Deccan. There are other tribes, such as the Rabari, or rearer of camels, who will be described with the classes appertaining to the whole desert.

Dhāt and Umrasūmra.

The Horrors of Humāyūn’s March.

We are now in the very region where Humayun suffered these miseries, and in its chief town, Umarkot, Akbar, the greatest monarch India ever knew, first saw the light. Let us throw aside the veil which conceals the history of the race of Humayun’s protector, and notwithstanding he is now but nominal sovereign of Umarkot, and lord [310] of the village of Chor,[^[14]] give him “a local habitation and a name,” even in the days of the Macedonian invader of India.

Dhāt.

Aror, Umarsūmra.

Aror.

We may follow Abu-l Fazl and Ferishta in their summaries of the history of ancient Sind, and these races. The former says: “In former times, there lived a Rāja named Siharas, whose capital was Alor. His sway extended eastward, as far as Kashmīr and towards the sea to Mekrān, while the sea confined it on the south and the mountains to the north. An invading army entered the country from Persia, in opposing which the Rāja lost his life. The invaders, contenting themselves with devastating part of the territory, returned. Rāē Sahi,[^[21]] the Rāja’s son, succeeded his father, by whose enlightened wisdom and the aid of his intelligent minister Rām, justice was universally administered and the repose of the country secured.... In the caliphate of Walīd bin Abdu’l Malik, when Hajjāj was governor of Irāk, he dispatched on his own authority Muhammad Kāsim, his cousin and son-in-law, to Sind, who fought Dāhir in several engagements.... After Muhammad Kāsim’s death, the sovereignty of this country devolved on the descendants of the Banu Tamīm Ansāri. They were succeeded by the Sūmrah race, who established their rule, and were followed by the Sammas, who asserted their descent from Jamshīd, and each of them assumed the name of Jām.”[^[22]]

Ferishta gives a similar version. “On the death of Mahomed Kasim, a tribe who trace their origin from the Ansarias established a government in Sind; after which the zamindars [lords of the soil or indigenous chiefs], denominated in their country Soomura, usurped the power, and held independent rule over the kingdom of Sinde for the space of five hundred years. These [312], the Soomuras, subverted the country of another dynasty called Soomuna [the Samma of Abu-l Fazl], whose chief assumed the title of Jam.”[^[23]]

The difficulty of establishing the identity of these tribes from the cacography of both the Greek and Persian writers, is well exemplified in another portion of Ferishta, treating of the same race, called by him Soomuna, and Samma by Abu-l Fazl. “The tribe of Sahna appears to be of obscure origin, and originally to have occupied the tract lying between Bekher and Tatta in Sinde, and pretend to trace their origin from Jemshid.” We can pardon his spelling for his exact location of the tribe, which, whether written Soomuna, Sehna, or Seemeh, is the Summa or Samma tribe of the great Yadu race, whose capital was Summa-ka-kot, or Sammanagari, converted into Minnagara, and its princes into Sambas, by the Greeks.[^[24]] Thus the Sodhas appear to have ruled at Aror and Bakhar, or Upper Sind, and the Sammas in the lower,[^[25]] when Alexander passed through this region. The Jarejas and Jams of Navanagar in Saurashtra claim descent from the Sammas, hence called elsewhere by Abu-l Fazl “the Sind-Samma dynasty”; but having been, from their amalgamation with the ‘faithful,’ put out of the pale of Hinduism, they desired to conceal their Samma-Yadu descent, which they abandoned for Jamshid, and Samma was converted into Jam.[^[26]]

We may, therefore, assume that a prince of the Sodha tribe held that division of the great Puar sovereignty, of which Aror, or the insular Bakhar, was the capital, when Alexander passed down the Indus: nor is it improbable that the army, styled Persian by Abu-l Fazl, which invaded Aror, and slew Raja Siharas, was a Graeco-Bactrian army led by Apollodotus, or Menander, who traversed this region, “ruled by Sigertides” (qu. Raja Siharas?) even to “the country of the Σῶρα,” or Saurashtra,[^[27]] where, according to their historian, their medals were existent when he wrote in the second century.[^[28]] The histories so largely quoted give us decided proof that Dahir, and his son [313] Raesa, the victims of the first Islamite invasion led by Kasim, were of the same lineage as Raja Siharas; and the Bhatti annals prove to demonstration, that at this, the very period of their settling in the desert, the Sodha tribe was paramount (see p. [1185]); which, together with the strong analogies in names of places and princes, affords a very reasonable ground for the conclusion we have come to, that the Sodha tribe of Puar race was in possession of Upper Sind, when the Macedonian passed down the stream; and that, amidst all the vicissitudes of fortune, it has continued (contesting possession with its ancient Yadu antagonist, the Samma) to maintain some portion of its ancient sovereignty unto these days. Of this portion we shall now instruct the reader, after hazarding a passing remark on the almost miraculous tenacity which has preserved this race in its desert abode during a period of at least two thousand two hundred years,[^[29]] bidding defiance to foreign foes, whether Greek, Bactrian, or Muhammadan, and even to those visitations of nature, famines, pestilence, and earthquakes, which have periodically swept over the land, and at length rendered it the scene of desolation it now presents; for in this desert, as in that of Egypt, tradition records that its increase has been and still is progressive, as well in the valley of the Indus as towards the Jumna.

Umarkot.

The Fate of the Sodha Tribe. Assassination of Mīr Bijar.

This episode, which properly belongs to the history of Marwar, or to Sind, is introduced for the purpose of showing the influence of the latter on the destinies of the Sodha princes. It was by Bijar, who fell by the emissaries of Bijai Singh, that the Sodha Raja was driven from Umarkot, the possession of which brought the Sindis into immediate collision with the Bhattis and Rathors. But on his assassination and the defeat of the Sind army on the Rann, Bijai Singh reinducted the Sodha prince to his gaddi of Umarkot; not, however, long to retain it, for on the invasion from Kandahar, this poor country underwent a general massacre and pillage by the Afghans, and Umarkot was assaulted and taken. When Fateh Ali made head against the army of Kandahar, which he was enabled to defeat, partly by the aid of the Rathors, he relinquished, as the price of this aid, the claims of Sind upon Umarkot, of which Bijai Singh took possession, and on whose battlements the flag of the Rathors waved until the last civil war, when the Sindis expelled them. Had Raja Man known how to profit by the general desire of his chiefs to redeem this distant possession, he might have got rid of some of the unquiet spirits by other means than those which have brought infamy on his name.

Chor.

The Sodha, and the Jareja, are the connecting links between the Hindu and the Muslim; for although the farther west we go the greater is the laxity of Rajput prejudice, yet to something more than mere locality must be attributed the denationalized sentiment which allows the Sodha to intermarry with a Sindi: this cause is hunger; and there are few zealots who will deny that its influence is more potent than the laws of Manu. Every third year brings famine, and those who have not stored up against it fly to their neighbours, and chiefly to the valley of the Indus. The [317] connexions they then form often end in the union of their daughters with their protectors; but they still so far adhere to ancient usage as never to receive back into the family caste a female so allied.[^[34]] The present Rana of the Sodhas has set the example, by giving daughters to Mir Ghulam Ali and Mir Sohrab, and even to the Khosa chief of Dadar; and in consequence, his brother princes of Jaisalmer, Bah and Parkar, though they will accept a Sodha princess to wife (because they can depend on the purity of her blood), yet will not bestow a daughter on the Rana, whose offspring might perhaps grace the harem of a Baloch. But the Rathors of Marwar will neither give to nor receive daughters of Dhat. The females of this desert region, being reputed very handsome, have become almost an article of matrimonial traffic; and it is asserted, that if a Sindi hears of the beauty of a Dhatiani, he sends to her father as much grain as he deems an equivalent, and is seldom refused her hand. We shall not here further touch on the manners or other peculiarities of the Sodha tribe, though we may revert to them in the general outline of the tribes, with which we shall conclude the sketch of the Indian desert.

Tribes.

Of the Muhammadan there are but two, Kalhora and Sahariya, concerning whose origin any doubt exists, and all those we are about to specify are Nayyads,[^[36]] or proselytes chiefly from Rajput or other Hindu tribes:

Zjat; Rajar; Umra; Sumra; Mair, or Mer; Mor, or Mohor; Baloch; Lumria, or Luka; Samaicha; Mangalia; Bagria; Dahya; Johya; Kairui; Jangaria; Undar; Berawi; Bawari; Tawari; Charandia; Khosa; Sadani; Lohanas.

The Nayyāds.

On the Bhattis, the Rathors, the Chauhans, and their offset the Mallani, we have sufficiently expatiated, and likewise on the Sodha; but a few peculiarities of this latter tribe remain to be noticed.

The Sodha Tribe.

The Kaurava Tribe.

The Dhāti Tribe.

The Lohāna Tribe.

The Arora Tribe.

The Bhātia Tribe.

Brāhmans.

The Rabāri Tribe.

Jat Tribes.

We shall now leave this general account of the Hindu tribes, who throughout Sind are subservient to the will of the Muhammadan, who is remarkable, as before observed, for intolerance. The Hindu is always second: at the well, he must wait patiently until his tyrant has filled his vessel; or if, in cooking his dinner, a Muslim should require fire, it must be given forthwith, or the shoe would be applied to the Hindu’s head.

The Sahariya Tribe.

The Khosa Tribe.

The Samaicha Tribe.

The Rājar Tribe.

The Umar Sūmra Tribe.

The Kalhora, Tālpuri Tribes.

Nūmri, Lūmri, or Lūka Tribe.

The Zott[^[54]] or Jat Tribe.—This very original race, far more numerous than perhaps all the Rajput tribes put together, still retains its ancient appellation throughout the whole of Sind, from the sea to Daudputra, but there are few or none in the thal. Their habits differ little from those who surround them. They are amongst the oldest converts to Islam.

The Mer, Mair Tribe.

The Mor, Mohor Tribe.

The Tāwari, Thori, or Tori Tribe.

Johya, Dahya, Mangalia Tribes.

Dāūdputra, Bahāwalpur State.

Daud Khan, the founder of Daudputra, was a native of Shikarpur, west of the Indus, where he acquired too much power for a subject, and consequently drew upon himself the arms of his sovereign of Kandahar. Unable to cope with them, he abandoned his native place, passed his family and effects across the Indus, and followed them into the desert. The royal forces pursued, and coming up with him at Sutiala, Daud had no alternative but to surrender, or destroy the families who impeded his flight or defence. He acted the Rajput, and faced his foes; who, appalled at this desperate act, deemed it unwise to attack him, and retreated. Daud Khan, with his adherents, then settled in the kachhi, or flats of Sind, and gradually extended his authority into the thal. He was succeeded by Mubarik Khan; he, by his nephew Bahawal Khan, whose son is Sadik Muhammad Khan, the present lord of Bahawalpur, or Daudputra, a name applied both to the country and to its possessors, “the children of David.”[^[58]] It was Mubarik who deprived the Bhattis of the district called Khadal, so often mentioned in the Annals of Jaisalmer, and whose chief town is Derawar, founded by Rawal Deoraj in the eighth century; and where the successor of Daud established his abode. Derawar was at that time inhabited by a branch of the Bhattis, broken off at a very early period, its chief holding the title of Rawal, and whose family since their expulsion have resided at Ghariala, belonging to Bikaner, on [325] an allowance of five rupees a day, granted by the conqueror. The capital of the “sons of David” was removed to the south bank of the Gara by Bahawal Khan (who gave it his name), to the site of an old Bhatti city, whose name I could not learn. About thirty years ago[^[59]] an army from Kandahar invaded Daudputra, invested and took Derawar, and compelled Bahawal Khan to seek protection with the Bhattis at Bikampur. A negotiation for its restoration took place, and he once more pledged his submission to the Abdali king, and having sent his son Mubarik Khan as a hostage and guarantee for the liquidation of the imposition, the army withdrew. Mubarik continued three years at Kabul, and was at length restored to liberty and made Khan of Bahawalpur, on attempting which he was imprisoned by his father, and confined in the fortress of Khangarh, where he remained nearly until Bahawal Khan’s death. A short time previous to this, the principal chiefs of Daudputra, namely, Badera Khairani, chief of Mozgarh, Khudabakhsh of Traihara, Ikhtiyar Khan of Garhi, and Haji Khan of Uchh, released Mubarik Khan from Khangarh and they had reached Murara, when tidings arrived of the death of Bahawal Khan. He continued his route to the capital; but Nasir Khan, son of Alam Khan, Gurgecha (Baloch), having formerly injured him and dreading punishment, had him assassinated, and placed his brother, the present chief, Sadik Muhammad, on the masnad: who immediately shut up his nephews, the sons of Mubarik, together with his younger brothers, in the fortress of Derawar. They escaped, raised a force of Rajputs and Purbias, and seized upon Derawar; but Sadik escaladed it, the Purbias made no defence [326], and both his brothers and one nephew were slain. The other nephew got over the wall, but was seized by a neighbouring chief, surrendered, and slain; and it is conjectured the whole was a plot of Sadik Khan to afford a pretext for their death. Nasir Khan, by whose instigation he obtained the masnad, was also put to death, being too powerful for a subject. But the Khairani lords have always been plotting against their liege; an instance of which has been given in the Annals of Bikaner, when Traihara and Mozgarh were confiscated, and the chiefs sent to the castle of Khangarh, the State prison of Daudputra. Garhi still belongs to Abdulla, son of Haji Khan, but no territory is annexed to it. Sadik Muhammad has not the reputation of his father, whom Bijai Singh, of Marwar, used to style his brother. The Daudputras are much at variance amongst each other, and detested by the Bhattis, from whom they have hitherto exacted a tribute to abstain from plunder. The fear of Kandahar no longer exists at Bahawalpur, whose chief is on good terms with his neighbour of Upper Sind, though he is often alarmed by the threats of Ranjit Singh of Lahore, who asserts supremacy over “the children of David.”

Diseases.

They have the usual infantine and adult diseases, as in the rest of India. Of these the sitala, or ‘smallpox,’ and the tijari, or ‘tertian,’ are the most common. For the first, they merely recommend the little patient to Sitala Mata; and treat the other with astringents in which infusion of the rind of the pomegranate is always (when procurable) an ingredient. The rich, as in other countries, are under the dominion of empirics, who entail worse diseases by administering mineral poisons, of whose effects they are ignorant. Enlargement of the spleen under the influence of these fevers is very common, and its cure is mostly the actual cautery.

Famines.

Productions, Animal and Vegetable.

The Wild Ass.

Rojh or Nilgae, Lions, etc.

Domestic Animals.

Flocks (here termed chang) of goats and sheep are pastured in vast numbers in the desert. It is asserted that the goat can subsist without water from the month of Karttik to the middle of Chait, the autumnal to the spring equinox [329]—apparently an impossibility: though it is well known that they can dispense with it during six weeks when the grasses are abundant. In the thals of Daudputra and Bhattipo, they remove to the flats of Sind in the commencement of the hot weather. The shepherds, like their flocks, go without water, but find a substitute in the chhachh, or buttermilk, after extracting the butter, which is made into ghi, and exchanged for grain, or other necessaries. Those who pasture camels also live entirely upon their milk, and the wild fruits, scarcely ever tasting bread.

Shrubs and Fruits.

The karil, or khair (the capparis, or caper-bush), is well known both in Hindustan and the desert: there they use it as a pickle, but here it is stored up as a culinary article of importance. The bush is from ten to fifteen feet in height, spreading very wide; there are no leaves on its evergreen twig-like branches, which bear a red flower, and the fruit is about the size of a large black currant. When gathered, it is steeped for twenty-four hours in water, which is then poured off, and it undergoes, afterwards, two similar operations, when the deleterious properties are carried off; they are then boiled and eaten with a little salt, or by those who can afford it, dressed in ghi and eaten with bread. Many families possess a stock of twenty maunds.

The sajji is a low, bushy plant, chiefly produced in the northern desert, and most abundant in those tracts of Jaisalmer called Khadal, now subject to Daudputra. From Pugal to Derawar, and thence by Muridkot, Ikhtyar Khan-ki-garhi, to Khairpur (Dair Ali), is one extensive thal, or desert, in which there are very considerable tracts of low, hard flat, termed chittram,[^[66]] formed by the lodgment of water [330] after rain, and in these spots only is the sajji plant produced. The salt, which is a sub-carbonate of soda, is obtained by incineration, and the process is as follows: Pits are excavated and filled with the plant, which, when fired, exudes a liquid substance that falls to the bottom. While burning, they agitate the mass with long poles, or throw on sand if it burns too rapidly. When the virtue of the plant is extracted, the pit is covered with sand, and left for three days to cool; the alkali is then taken out, and freed from its impurities by some process. The purer product is sold at a rupee the ser (two pounds weight); of the other upwards of forty sers are sold for a rupee. Both Rajputs and Muhammadans pursue this employment, and pay a duty to the lord paramount of a copper pice on every rupee’s worth they sell. Charans and others from the towns of Marwar purchase and transport this salt to the different marts, whence it is distributed over all parts of India. It is a considerable article of commerce with Sind, and entire caravans of it are carried to Bakhar, Tatta, and Cutch. The virtue of the soda is well understood in culinary purposes, a little sajji added to the hard water soon softening the mess of pulse and rice preparing for their meals; and the tobacconists use considerable quantities in their trade, as it is said to have the power of restoring the lost virtues of the plant.

Grasses.

Melons.

ITINERARY[^[69]]

Jaisalmer to Sehwan, on the right bank of the Indus, and Haidarabad, and return by Umarkot to Jaisalmer

Kuldra (5 coss).—A village inhabited by Paliwal Brahmans; two hundred houses; wells.

Gajia-ki-basti (2 do.).—Sixty houses; chiefly Brahmans; wells.

Khaba (3 do.).—Three hundred houses; chiefly Brahmans; a small fort of four bastions on low hills, having a garrison of Jaisalmer.

—An assemblage of hamlets of four or five huts on one spot, about a mile distant from each other, conjointly called Sum, having a burj or tower for defence, garrisoned from Jaisalmer; several large wells, termed beria; inhabitants, chiefly Sindis of various tribes, pasture their flocks, and bring salt and khara (natron) from Deo Chandeswar, the latter used as a mordant in fixing colours, exported to all parts. Half-way between Sum and Mulana is the boundary of Jaisalmer and Sind.

Mulana[^[70]] (24 coss).—A hamlet of ten huts; chiefly Sindis; situated amidst lofty sandhills. From Sum, the first half of the journey is over alternate sandhills, rocky ridges (termed magra), and occasionally plain; for the next three, rocky ridges and sandhills without any flats, and the remaining nine coss a succession of lofty tibas. In all this space of twenty-four coss there are no wells, nor is a drop of water to be had but after rain, when it collects in some old tanks or reservoirs, called nadi and taba, situated half-way, where in past times there was a town.

It is asserted, that before the Muhammadans conquered Sind and these regions, the valley and desert belonged to Rajput princes of the Pramar and Solanki tribes; that the whole thal (desert) was more or less inhabited, and the remains of old tanks and temples, notwithstanding the drifting of the sands, attest the fact. Tradition records a famine of twelve years’ duration during the time of Lakha Phulani, in the twelfth century, which depopulated the country, when the survivors of the thal fled to the kachhi, or flats of the Sind. There are throughout still many oases or cultivated patches, designated by the local terms from the [332] indispensable element, water, which whether springs or rivulets, are called wah, bah, beria, rar, tar, prefixed by the tribe of those pasturing, whether Sodhas, Rajars, or Samaichas. The inhabitants of one hamlet will go as far as ten miles to cultivate a patch.

These are all hamlets of about ten huts, inhabited by Rajars, who cultivate patches of land or pasture their flocks of buffaloes, cows, camels, goats, amidst the thal; at each of these hamlets there are plenty of springs; at Rajar-ki-basti there is a pool called Mahadeo-ka-dah. (See p. [1263] above.)

Deo Chandeswar Mahadeo (2 do.).—When the Sodha princes held sway in these regions, there was a town here, and a temple to Mahadeo, the ruins of which still exist, erected over a spring called Suraj kund, or fountain of the Sun. The Islamite destroyed the temple, and changed the name of the spring to Dinbawa, or ‘waters of the faith.’ The kund is small, faced with brick, and has its margin planted with date trees and pomegranates, and a Mulla, or priest from Sind, resides there and receives tribute from the faithful. For twelve coss around this spot there are numerous springs of water, where the Rajars find pasture for their flocks, and patches to cultivate. Their huts are conical like the wigwams of the African, and formed by stakes tied at the apex and covered with grass and leaves, and often but a large blanket of camel’s hair stretched on stakes.

Chandia-ki-basti (2 coss).—Hamlet inhabited by Muslims of the Chandia tribe, mendicants who subsist on the charity of the traveller.

Rajar-ki-basti Samaicha-ki-do		(2 do.). (2 do.).
Rajar Do. Do. Do. Do. Do.	do. do. do. do. do. do.	(2 do.). (2 do.). (2 do.). (2 do.). (2 do.). (2 do.).

Purwas, or hamlets of shepherds, Samaichas, Rajars, and others, who are all migratory, and shift with their flocks as they consume the pastures. There is plenty of water in this space for all their wants, chiefly springs.

Udhania (7 do.).—Twelve huts; no water between it and the last hamlet.

Nala (5 do.).—Descent from the thal or desert, which ceases a mile east of the nala or stream, said to be the same which issues from the Indus at Dara, above Rohri-Bakhar; thence it passes east of Sohrab’s Khairpur, and by Jinar to Bersia-ka-rar, whence there is a canal cut to Umarkot and Chor.

Mitrao (4 do.).—Village of sixty houses, inhabited by Baloch; a thana, or post here from Haidarabad; occasional low sandhills.

Mir-ki-kui (6 do.).—Three detached hamlets of ten huts each, inhabited by Aroras.

Sheopuri (3 do.).—One hundred and twenty houses, chiefly Aroras: small fort of six bastions to the south-east, garrisoned from Haidarabad.

Kamera-ka-Nala (6 do.).—This nala issues from the Indus between Kakar-ki-basti and Sakrand, and passes eastward; probably the bed of an old canal, with which the country is everywhere intersected.

Sakrand (2 do.).—One hundred houses, one-third of which are Hindus; patches of cultivation; numerous watercourses neglected; everywhere overgrown with jungle, chiefly jhau and [333] khejra (tamarisk and acacia). Cotton, indigo, rice, wheat, barley, peas, grain, and maize grow on the banks of the watercourses.

Jatui (2 do.).—Sixty houses; a nala between it and Jatui.

Kazi-ka-Shahr (4 do.).—Four hundred houses; two nalas intervene.

Makera (4 coss).—Sixty houses; a nala between it and Jatui.

Kakar-ki-basti (6 do.).—Sixteen houses; half-way the remains of an ancient fortress; three canals or nalas intervening; the village placed upon a mound four miles from the Indus, whose waters overflow it during the periodic monsoon.

Pura or Hamlet (1 do.).—A ferry.

The Indus (1 do.).—Took boat and crossed to

Sewan or Sehwan (1½ do.).—A town of twelve hundred houses on the right bank, belonging to Haidarabad[^[71]] [334].

Sehwan to Haidarabad

Jat-ki-basti (2 coss).—The word jāt or jat is here pronounced Zjat. This hamlet ‘basti,’ is of thirty huts, half a mile from the Indus: hills close to the village.

Samaicha-ki-basti (2½ coss).—Small village.

Lakhi (2½ do.).—Sixty houses; one mile and a half from the river: canal on the north side of the village; banks well cultivated. In the hills, two miles west, is a spot sacred to Parbati and Mahadeo, where are several springs, three of which are hot.[^[72]]

Umri (2 do.).—Twenty-five houses, half a mile from River; the hills not lofty, a coss west.

Sumri (3 do.).—Fifty houses, on the River hills; one and a half coss west.

Sindu or San (4 do.).—Two hundred houses and a bazar, two hundred yards from the River; hills one and a half coss west.

Manjhand (4½ do.).—On the River two hundred and fifty houses, considerable trade; hills two coss west.

Umar-ki-basti (3 do.).—A few huts, near the river.

Sayyid-ki-basti (3 do.).

Shikarpur (4 do.).—On the river; crossed to the east side.

Haidarabad (3 coss).—One and a half coss from the river Indus. Haidarabad to Nasarpur, nine coss; to Sheodadpur, eleven do.; to Sheopuri, seventeen do.; to Rohri-Bakhar, six do.—total forty-three coss.

Haidarabad via Umarkot, to Jaisalmer

Sindu Khan ki-basti (3 do.).—West bank of Phuleli river.

Tajpur (3 do.).—Large town, north-east of Haidarabad [335].

Katrel (1½ do.).—A hundred houses.

Nasarpur (1½ do.).—East of Tajpur, large town.

Alahyar-ka-Tanda (4 do.).—A considerable town built by Alahyar Khan, brother of the late Ghulam Ali, and lying south-east of Nasarpur. Two coss north of the town is the Sangra Nala or Bawa,[^[73]] said to issue from the Indus between Hala and Sakrand and passing Jandila.

Mirbah (5 do.).—Forty houses; Bah, Tanda, Got, Purwa, are all synonymous terms for habitations of various degrees.

Sunaria (7 do.).—Forty houses.

Dangana (4 do.).—To this hamlet extend the flats of Sind. Sandhills five and six miles distant to the north. A small river runs under Dangana.

Karsana (7 do.).—A hundred houses. Two coss east of Karsana are the remains of an ancient city; brick buildings still remaining, with well and reservoirs. Sandhills two to three coss to the northward.

Umarkot (8 do.).—There is one continued plain from Haidarabad to Umarkot, which is built on the low ground at the very extremity of the thal or sand-hills of the desert, here commencing. In all this space, estimated at forty-four kachha coss, or almost seventy miles of horizontal distance, as far as Sunaria the soil is excellent, and plentifully irrigated by bawahs, or canals from the Indus. Around the villages there is considerable cultivation; but notwithstanding the natural fertility, there is a vast quantity of jungle, chiefly babul (Mimosa arabica), the evergreen thal, and thal or tamarisk. From Sunaria to Umarkot is one continued jungle, in which there are a few cultivated patches dependent on the heavens for irrigation; the soil is not so good as the first portion of the route.

Katar (4 do.).—A mile east of Umarkot commences the thal or sandhills, the ascent a hundred and fifty to two hundred feet. A few huts of Samaichas who pasture; two wells.

Dhat-ki-basti (4 do.).—A few huts; one well; Dhats, Sodhas, and Sindis cultivate and pasture.

Dharnas (8 coss).—A hundred houses, chiefly Pokharna Brahmans and Banias, who purchase up the thal from the pastoral tribes, which they export to Bhuj and the valley. It is also an entrepôt for trade; caravans from the east exchange their goods for the thal, here very cheap, from the vast flocks pastured in the Rui.

Kherlu-ka-Par (3 do.).—Numerous springs (thal) and hamlets scattered throughout this tract.

Lanela (1½ do.).—A hundred houses; water brackish; conveyed by camels from Kherlu.

Bhoj-ka-Par (3 do.).—Huts; wells; patches of cultivation.

Bhu (6 do.).—Huts.

Garara (10 do.).—A small town of three hundred houses, belonging to Sawai Singh Sodha, with several thal or hamlets attached to it. This is the boundary between Dhat or the Sodha raj and Jaisalmer. Dhat is now entirely incorporated in Sind. A thal, or collector of the transit duties, resides here.

Harsani (10 do.).—Three hundred houses, chiefly Bhattis. It belongs to a Rajput of this tribe, now dependent on Marwar [336].

Jinjiniali (10 do.).—Three hundred houses. This is the fief of the chief noble of Jaisalmer; his name Ketsi,[^[74]] Bhatti. It is the border town of Jaisalmer. There is a small mud fortress, and several talaos, or sheets of water, which contain water often during three-fourths of the year; and considerable cultivation in the little valleys formed by the thal, or sand-ridges. About two miles north of Jinjiniali there is a village of Charans.

Gaj Singh-ki-basti (2 do.).—Thirty-five houses. Water scarce, brought on camels from the Charan village.

Hamirdeora (5 do.).—Two hundred houses. There are several thal or pools, about a mile north, whither water is brought on camels, that in the village being saline. The ridge of rocks from Jaisalmer here terminates.

Chelak (5 do.).—Eighty houses; wells; Chelak on the ridge.

Bhopa (7 do.).—Forty houses; wells; small thal or pool.

Bhao (2 do.).—Two hundred houses; pool to the west; small wells.

Jaisalmer (5 do.).—Eighty-five and a half coss from Umarkot to Jaisalmer by this route, which is circuitous. That by Jinjiniali 26 coss, Girab 7, Nilwa 12, Umarkot 25—in all 70 pakka coss, or about 150 miles. Caravans or kitars of camels pass in four days, kasids or messengers in three and a half, travelling night and day. The last 25 coss, or 50 miles, is entire desert: add to this 44 short coss from Haidarabad to Umarkot, making a total of 129½ coss. The most direct road is estimated at 105 pakka coss, which, allowing for sinuosities, is equal to about 195 English miles.

Total of this route, 85½ coss.

Jaisalmer to Haidarabad, by Baisnau

Kuldar (5 coss).

Khaba (5 do.).

Lakha-ka-ganw (30 do.).—Desert the whole way; no hamlets or water.

Baisnau (8 do.).

Bersia-ka-Rar (16 do.).—Wells.

Thipra (3 do.).

Mata-ka-dher (7 do.).—Umarkot distant 20 coss.

Jandila (8 do.).

Alahyar-ka-Tanda[Alahyar-ka-Tanda] (10 do.).—Sankra, or Sangra thal.

Total of this route, 103 coss.

Jaisalmer, by Shahgarh, to Khairpur of Mir Sohrab

Anasagar (2 do.).

Chonda (2 do.).

Pani-ka-tar (3 do.).—Tar or Tir, springs [337].

Pani-ki-kuchri (7 do.).—No village.

Kuriala (4 do.).

Shahgarh (20 do.[^[75]]).—Rui or waste all this distance. Shahgarh is the boundary; it has a small castle of six bastions, a post of Mir Sohrab, governor of Upper Sind.

Garsia (6 do.).

Garhar (28 do.).—Rui or desert the whole way; not a drop of water. There are two routes branching off from Garhar, one to Khairpur, the other to Ranipur.

Nala (2 do.).—The same stream which flows from Dara, and through the ancient city of Alor; it marks the boundary of the desert.

Khairpur[^[76]] (18 coss).—Mir Sohrab, governor of Upper Sind, and brother of the prince of Haidarabad, resides here. He has erected a stone fortress of twelve bastions, called Nawakot or New-castle. The 18 coss from the thal to Khairpur is flat, and marks the breadth of the valley here. The following towns are of consequence.

Khairpur to Larkhana.—Twenty coss west of the Indus, held by Karam Ali, son of the prince of Haidarabad.

Khairpur to Lakhi.—Fifteen coss, and five from Shikarpur.

Khairpur to Shikarpur (20 do.).

Garhar to Ranipur

Pharara (10 do.).—A village of fifty houses, inhabited by Sindis and Karars; several hamlets around. A dani, or collector of transit dues, resides here on the part of Mir Sohrab, the route being travelled by kitars or caravans of camels. The nala from Dara passes two coss east of Pharara, which is on the extremity of the desert. Commencement of the ridge called Takar, five coss west of Pharara, extending to Rohri Bakhar, sixteen coss distant from Pharara. From Pharara to the Indus, eighteen coss, or thirty miles breadth of the valley here.

Ranipur[^[77]] (18 do.).

Jaisalmer to Rohri Bakhar

Kuriala (18 do.).—See last route.

Banda (4 do.).—A tribe of Muslims, called Undar, dwell here.

Gotru (16 do.).—Boundary of Jaisalmer and Upper Sind. A small castle and garrison of Mir Sohrab’s; two wells, one inside; and a hamlet of thirty huts of Samaichas and Undars; thal heavy.

Udat (32 do.).—Thirty huts of shepherds; a small mud fortress. Rui, a deep and entire desert, throughout all this space; no water [338].

Sankram or Sangram (16 do.).—Half the distance sand-hills, the rest numerous temporary hamlets constructed of the thal, or maize stalks; several water-courses.

Nala-Sangra (½ do.).—This nala or stream is from Dara, on the Sind, two coss and a half north of Rohri Bakhar; much cultivation; extremity of the sand-hills.

Targatia (½ do.).—A large town; Bankers and Banias, here termed Karar and Samaichas.

Low ridge of hills, called Takar (4 do.).—This little chain of silicious rocks runs north and south; Nawakot, the Newcastle of Sohrab, is at the foot of them; they extend beyond Pharara, which is sixteen coss from Rohri Bakhar. Gumat is six coss from Nawakot.

On the ridge, on the left bank of the Indus. Crossed over to Bakhar; breadth of the river near a mile. Bakhar is an island, and the other branch to Sakhar is almost a mile over also. This insulated rock is of silex, specimens of which I possess. There are the remains of the ancient fortress of Mansura, named in honour of the Caliph Al-Mansur, whose lieutenants made it the capital of Sind on the opening of their conquests. It is yet more famed as the capital of the Sogdoi of Alexander; in all probability a corruption of Sodha, the name of the tribe which has ruled from immemorial ages, and who till very lately held Umarkot.

N.B.—Kasids or messengers engage to carry despatches from Jaisalmer to Rohri Bakhar in four days and a half; a distance of one hundred and twelve coss.

Bakhar to Shikarpur