Pramāṇa · Search · About · API

Explorations in Sind (MASI 48) c5.s1: Section 1.—In the Delta Country

Explorations in Sind (MASI 48) c5.s1

Language
English

Section 1.—In the Delta Country

c5.s1.1

The First Journey.—After the lapse of a year since the excavations of Jhukar, I was deputed to carry out an exploratory survey of the Lower Indus Valley. The purpose for which this work had to be undertaken was twofold. In the first place, our knowledge of the southern limit of the Indus culture was extremely meagre, and it became necessary therefore to gather as much information as possible regarding ‘mounds’ situated in the Delta country. Secondly, a list of sites definitely ascribed to the Prehistoric period by R. D. Banerji was in the hands of Sir John Marshall, which required a thorough checking in the light of observations made on the spot. When all necessary preliminaries for this survey were settled in correspondence with the Commissioner in Sind, and tents and other equipage obtained, we left Simla at the end of October, 1929, and proceeded directly to Karachi which was fixed as the starting point of our journey.

c5.s1.2

Bhambor.—Our first camp after leaving Karachi was Gharo which lies in the midst of a sandy plain near Dabheji Station on the Kotri-Karachi Line of the North Western Railway. About 3½ miles from this place is the rocky site of Bhambor on the Gharo Creek. It stands on a hill about fifty feet high, locally known as ‘Sasui-jo-takar,’¹ or ‘the hill of Sasui,’ being the scene of the legend² of Sasui and her lover Punhu which is so familiar to the people of Sind. According to some scholars the site represents an outpost guarding the Gharo Creek and is not earlier than the Arab invasion of 712 A.D.³ This dating was confirmed by the surface finds of glazed pottery, a kind of ‘honey-comb’ terracotta and small copper coins of the Arab rulers of Sind. The terracotta has been noticed also by Cousens,⁴ who thinks that it was used as moulds for coins. But in all probability they are fragments of that peculiar hand-mill known as jandi in Sind. Of these many specimens were subsequently collected from various sites of the Muhammadan period during my journey. Exploratory trenches were sunk at Bhambor to find out if it rested on pre-Muhammadan deposits, but this expectation was not fulfilled. The buildings, which were made of stone, probably belonged to a single period of occupation. By the side of a stone wall, revealed in one of the trenches, about eighty copper coins of the Arab rulers were collected together with fragments of glazed pottery, like those noticed on the surface of the mound.

c5.s1.3

¹ This piece belongs to a lot of thirteen coins. The rest are of the type of Pl. XIV, figs. 11, 12.

c5.s1.4

Tatta and Budhjo Takar.—Further work at Bhambor being thought unnecessary, the camp was shifted to Tatta, about twenty-four miles to the east of Gharo. In centuries past Tatta was the capital of Sind, and had the reputation of being one of the biggest marts of India. On a spur of the Makli hills are the ruins of Kalankot, about five miles to the southwest of Tatta. It is a fort of the Muhammadan times with a mosque on its west, and is surrounded on three sides by a lake, representing no doubt an arm of the Indus which once used to flow by the side of Tatta. An Englishman writing of Tatta in the 17th century observes, “the town is so encompassed with the river Indus that it makes a peninsula.”⁵ There was nothing at Tatta or at Kalankot to show that this area had ever been occupied in an earlier period. From Tatta was visited Budhjo Takar, a couple of hillocks on the left bank of the Indus. This is just opposite Jherruck, thirteen miles from Meting station which was reached by train from Jungshahi, the nearest station to Tatta. The road from Tatta to Jungshahi is an excellent one for Sind. This road as well as the one from Meting to Jherruck was found motorable, so that we could reach Jherruck in the course of a few hours. But crossing the Indus to Budhjo Takar and the return journey to Jherruck took much time, as opposite the latter place there was insufficient water and our boat had to follow an extremely circuitous route to reach the other bank. Eventually, the boat had to be left a quarter of a mile away from the bank, and this distance we had to wade through mud and shallow water. Towards the close of the day we reached Budhjo Takar. At this place, Mr. G. E. L. Carter

c5.s1.5

¹ Banerji, Prog. Rep., A. S. I., W. C., 1919, p. 57. ² Burton, Sindh (1851), pp. 389 and 92 ff. Cf. also Elliot, History of India, Vol. I, pp. 332-36. ³ Cousens, Antiquities of Sind, pp. 80-81. ⁴ Ibid. p. 80 and Pl. XV. ⁵ Raverty, J. A. S. B., 1892, Part I, p. 321.

c5.s1.6

E

c5.s1.7

is said to have collected some flints ‘upon two tumuli of brick débris.’¹ The tops of the hillocks no doubt looked like tumuli from a distance, but in reality they were nothing but the outcrop of natural rock and did not show any traces of ruins upon them. Not a single flint flake or potsherd could be detected in the whole area, and it did not seem to have ever been a site for human habitation. A local contractor had engaged labourers to cut away stone from various parts of the hillocks, and when we were on the spot they were still busy in collecting their material. Being questioned they told us that they had been for some time past working in this area, but never saw any structural remains, pottery or chert anywhere. It is therefore not clear where exactly Carter got his flints. In the Indian Museum, Calcutta,² there is a flint ‘scraper’ stated to have been found by Dr. Fedden of the Geological Survey, on the surface at ‘Jhirak in Lower Sind,’ in 1876. The exact findspot of this scraper also we have not been able to locate. It is quite unlike the implements peculiar to the Indus civilization, and has probably to be regarded as a relic of the Neolithic times.

c5.s1.8

The Tharro Hill near Gujo.—Returning from Jherruck we proceeded to Gujo, midway between Tatta and Gharo. The village of Gujo lies about ten miles to the west of Tatta on the Karachi road. This place had also been previously visited by Carter, about which Cousens writes: “At the Tharro near Gujo.....where there is a wonderful neolithic city, he made a fine collection of flints.”² Tharro is a small hill (Pl. IV, a) about two miles to the southwest of Gujo. The hill rises thirty to thirty-five feet above the plain and has a more or less flat top. Its northern side is occupied by the remains of a series of Muhammadan tombs and a temple, locally known as the Hāt,³ which has decayed almost beyond recognition. To the south of these is an area extending to about 1,600 feet by 800 feet, which is studded with innumerable chert flakes. There is a range of low mounds or heaps of earth, five to six feet in height, enclosing portions of this area. This probably led Carter to think that it was the site of ‘a walled city.’ At three places on the rocky surface of this hill are carved a number of symbols. At one of these spots no less than six signs have been engraved, but at the other two only a single sign is found. Among these ‘the man,’ ‘the bow and arrow’ and ‘the intersected rectangle’ are reminiscent of similar ones occurring on the seals of Harappa and Mohenjo-daro, although the available evidence is not enough to establish a prehistoric date for these symbols or trace their connection with the pictographic script of the Indus Valley.

c5.s1.9

Prehistoric Pottery.—Excavations across the chain of mounds mentioned above led to the discovery of a fairly large number of potsherds, mostly of a light red ware. The bulk of this pottery appears now quite plain, but some specimens bear traces to show that originally they had painted decoration. One of the painted pieces (Th 54) is illustrated in Pl. XVIII, 34. It is part of a vase ornamented with rectangles filled in by cross-lines. The pattern is executed in chocolate on the natural ground of the pottery, and above it there are traces of a reddish brown band. This fragment agrees, in point of technique and design, with certain potsherds from Amri shown in Pl. XVIII. Another interesting sherd is Th 53 (Pl. XVIII, 33) which represents a class of buff ware with buff slip. It belonged to a vase having a chocolate band at neck and some uncertain ornament painted in this same colour below. Th 53, and another fragment Th 23, are shown in their profile view in Pl. XXXVIII, 24, 17. Although it is difficult to judge of the actual shape of the pots of which they form part, it is clear that they conform to a type of straight-shouldered vessels.

c5.s1.10

¹ Antiquities of Sind, pp. 46-47. Cf. also Carter in Ind. Ant., 1932, p. 88, n. 13. ² Coggin Brown, Catalogue, p. 121, No. 816. The findspot is incorrectly given as ‘Shirak’ in the Catalogue. But in the original register of the Museum it appears as ‘Jhirak.’ ³ Carter describes it in Ind. Ant., 1932, p. 87 (see Pl. I, fig. 3 and Pl. II, fig. 4).

c5.s1.11

Contents of small mounds.—A feature that at once attracts notice, on this part of the hill, is a group of over one hundred small isolated mounds, or piles of stones, which at first sight appear to be Muhammadan graves. Since, however, they are oriented north to south in some cases and east to west in others, it was thought worth while to examine their contents. The majority of the mounds are rectangular, and the rest circular in shape. Three of the rectangular ones were opened, and after a quantity of earth and stones had been removed the remains of a chamber in each one of them, none exceeding nine feet in length and four feet in breadth, were laid bare. The chambers are crudely built; their walls which are of rough hewn stones stand only two to three feet high. From each of the chambers were recovered pottery articles comprising vases,¹ both large and small, and dishes-on-stand (Pl. XVI, 24: ht. 16·5"), the latter recalling the specimens discovered at the chalcolithic sites of the Indus Valley. No bones were found in the course of the digging, and it is difficult to imagine what purpose these deposits served. The pottery is unpainted red ware which does not furnish any definite criteria as to its age, although it is probable that it may, along with the rude stone structures with which it was found associated, belong to some prehistoric epoch, like the chert flakes and cores lying about on the top of the hill.

c5.s1.12

Nature of the Site.—Besides the pottery, a collection of flints numbering over two hundred and sixty was made from Tharro. These comprise flakes as well as cores (Pl. XIX, 19-22), some of the former being quite of the ‘pigmy’ type. There was no trace, however, of any dwellings in this area and the earth deposit over the natural rock was found to be only a few inches deep. In all likelihood Tharro hill was merely a centre of flint-knapping industry, and not a regular dwelling site. The future explorer may not, therefore, find it a suitable ground for excavation, but it may be useful to open a few more of the small mounds.

c5.s1.13

Buddhist Remains.—Some of these small mounds, it may be noted, contain deposits of a much later period. From one of the circular mounds, only when six inches of earth was removed, there came out a large clod of fine clay which, when broken, yielded over a hundred offertory tablets of the Buddhists. Two types could be distinguished among these tablets of unbaked clay: one figuring a seated Buddha and another a row of chaityas, in either case with the Buddhist creed written in North-Indian characters of the 7th-8th century A.D. The tablets are generally similar to those discovered in the excavations at Mirpurkhas² in the Thar and Parkar District.

c5.s1.14

¹ E. g. Th 7, ht. 16·5": Pl. XVI, 25. ² Cf. Antiquities of Sind, p. 91, fig. 11.

c5.s1.15

E 2

c5.s1.16

Remains of the Muhammadan period.—The Muhammadan tombs on the northern section of the hill deserve more than a passing mention, as some of them exhibit fine workmanship and probably represent a style of architecture earlier than that of the monuments on the Makli hills at Tatta. The most interesting monument on the Tharro hill, probably dating from the 14th century, is a rectangular enclosure, approximately 28 feet by 18 feet, within which there are two tombs. It is open to the sky and entered on the south by a door surmounted by a heavy lintel. The entire structure is of stone, and the enclosing wall, which is about 6 feet high, is crowned by a battlemented parapet. Below this parapet there are six courses of ashlar masonry. The door, the mihrab as well as the tomb are beautifully carved with geometric and arabesque patterns, but the rest of the building is severely plain, indicating a comparatively early age. Similar tombs, but of lesser importance, were also noticed near the Mausoleum of Sheikh Haji Turabi close by. Built into a wall of this building is an Arabic inscription which according to the Tufatu-l-Kiram is dated in 171 A. H.¹ A photograph of the inscription supplied to Mr. Ghulam Yazdani has, however, elicited the information that it is dated in 782 A. H., equivalent to 1380 A.D., and in the reign of Firoz Shah Tughlaq, and that it has nothing to do with the tomb itself or with the memory of the saint who is said to have been a contemporary of the Abbaside Khalifa Mansur. The Tufatu-l-Kiram mentions that the Sheikh ‘took the fort of Tharra, in the district of Sakura, the city of Baghar, Bhambur, and some other places in western Sind.’² There is no doubt the ‘Tharra’ is the same as the Tharro hill, and ‘Sakura’ identical with the Taluka now called Mirpur Sakro.

c5.s1.17

Across the Baghar.—Retracing our steps to Tatta, a trip by a motor car to some of the eastern Talukas of the Delta, namely Mirpur Batoro, Jati, Guni and Dehro Mohbat, was arranged. Accordingly, the Baghar river, which is a branch of the Indus, was crossed on the 22nd of November, and the same day we reached Mirpur Batoro. Our motor could be taken with us in the steamer that plies between the right bank of the Baghar and Syedpur on the other bank. From Mirpur Batoro the mounds at Shahkapur, Kakeja, Katbanan and Rarri were examined, but none of them appeared to be older than Muhammadan times. After inspecting a few more of these late mounds we came to Tando Muhammad Khan on the Fuleli canal. Near this place, on the spurs of the Gunjo Takar Range, there are two Buddhist stūpas, of which the most well known is Sudheranjo Daro. It is situated close to the village of Saidpur and the Gajah Wah, a branch of the Fuleli. This stūpa has been excavated and conserved by the Archæological Survey. Both the stūpas, however, are built on rocky foundations, and as such there was no possibility of discovering here any earlier strata. Next were examined the mounds at Nidamani and Tharri, which are situated between Matli and Tando Ghulam Ali, but none of them appeared to be even of the pre-Muhammadan period. Kotri was reached by the middle of December, and from there, as well as from Hyderabad on the opposite bank of the Indus, journeys were made in diverse directions. Mention may here be made only of the visits to Mirpurkhas, Umarkot and Badin.

c5.s1.18

¹ Cf. Karachi District Gazetteer, p. 100. ² Elliot, History of India, Vol. I, p. 446.

c5.s1.19

Mirpurkhas, Umarkot and Badin.—Near the town of Mirpurkhas, in the Thar and Parkar District, there is the site known as Kahujo-daro, where excavations were conducted by Cousens and Bhandarkar in 1909-10 and 1916-17 respectively. These resulted in the clearance of a Buddhist stūpa which has since been conserved by the Archæological Department. But encircling the stūpa there are several mounds still awaiting excavation, which probably mark the monastery area. In order to test whether earlier remains exist below the level of the stūpa, a trial trench was excavated to its north. This yielded, at a depth of five to six feet, a chalcedony bead, two copper coins and some plain pottery fragments. Digging was also carried on to the south of the stūpa, but it did not bring to light anything that could be assigned to the prehistoric period. While in Mirpurkhas several other mounds were visited in this District, including those near Shadipalli Station, at Khirah about thirteen miles to the northeast of Mirpurkhas and those near Sindhri about fourteen miles to the northeast of the same town, and the mound called Mumalji-mari near Umarkot. But at none of these places could be found any trace of the early culture. On return to Hyderabad we proceeded immediately to visit a few mounds in the Badin area. Near Peeru Lishari Station of the Hyderabad-Badin Line, we came across a late mound called Otjo-daro which is cut through by the Badin road. Half a mile to the south of Badin town were also examined the ruins of Old Badin, a cluster of tiny mounds with a varying height of six to eight feet. The site certainly belongs to the Muhammadan times, a fact to which the fragments of pottery handmills (jandi), glazed pottery and small copper coins lying here and there bear witness. Some of the other mounds visited near Badin are Kandarwaro Daro, about three miles north of the town and Dodojo-mari or Ropa-mari, about five miles to the southwest of Bahadmi which is twenty-two miles to the south of Badin. Dodojo-mari is situated in the midst of a barren sandy plain which beyond Badin takes the character of the Great Rann of Cutch. At none of these sites could there be seen anything of the pre-Muhammadan period.

c5.s1.20

Recent formation of the Delta.—The result of these journeys in the Delta is not altogether valueless, in spite of its negative character. It suggests that the present delta probably came into existence in comparatively recent times, and demonstrates also the futility of searching for early remains, beyond the hilly region of Lower Sind. As noticed by Raverty, the district of Badin was the southernmost part of Sind in the reign of Akbar, when the Sarkar of Tattah, which included Badin, was bounded on the south by the Ocean.¹ If a line is drawn, connecting Tatta and Gujo in the west with Badin in the east, that would in all likelihood represent the utmost limit to which the Indus country could have extended in the south in the prehistoric period.

c5.s1.21

¹ J. A. S. B., 1892, Part I, p. 317 ; Jarret, Ain-i-Akbari, Vol. II, p. 336.

Linked texts (10)