The army spent the season of the rains in Delhi and started for the south on the campaign against Ma’bar on November 20, 1310. A muster of the army was taken at Tankal, otherwise written Natgal, on the Jumna, and marching orders were given on the 2nd December following. In one straight march the army came to a place called Katihun in 21 days, and a further 17 days took them to Gurgaom, crossing three rivers of which the biggest was the Narmada. This summary description gives the impression that the route taken was the usual route in which the army met with no incidents of an untoward character. In all probability that route was the route of the return march from Warangal we described above. The army would have come down as far as Muttra, and turned south-westwards along the great highroad through Bharatpur, Savai-Madhopur, Kota, Jhalrapetan as far as Nagda. There is no place that is identifiable with anything like Katihun that Amīr Khusrau speaks of; at any rate, there is nothing satisfactory so far. But having regard to the length of march and the summary description, we may possibly take it that the stage Katihun was somewhere near the region of Nagda. Gurgaom must be located somewhere between Burhanpur and the Tapti, a distance of about five marches from where Deogiri actually is. The road taken this time probably went from the Narmada at Maheswar southwards to Pansemal, from which Nandurbar, Sindkheda and Dhulian and Jalgaom can all be reached across the Tapti; and beyond them lay Chalesgaom, and across the Sahyadri, Deogiri. There is nothing like Gurgaom on the maps in this locality, and we must therefore remain content with not being able to find its modern equivalent. Leaving Gurgaom, the army reached the Tapti, and crossing it reached Deogiri in five marches after leaving Gurgaom, on the 17th Ramzan, equivalent to 8th February 1311. There the army made a halt, and according to Barni, Ram Deo having died and Sankar Dev being the ruler, the malik appointed some one to attend to the needs of the army marching on the further journey. From there five marches took them across the three rivers, Sini, Godavari and Binhur to the place which Amīr Khusrau calls the Kharabābād of a Paras Deo Dalvi. Working on the account of Elliot alone, I took Kharābābad as an actual name of a place. But with the translation before me now it strikes me it is merely Amīr Khusrau’s description where he plays upon an attribute that he himself had given to Deogiri by calling it Almanābad, a city of safety, and by way of contrast, he calls the other Kharābābād, which he gives again to Kandur in the south. That must have been the headquarters of the southern government under the general Parsu Deo, the Dalavai, Parsu being the contracted form of Parasuram. The road starting from Deogiri and going southwards has to cross three rivers whichever way it went. But the usual way in those days seems to have been from Deogiri through the interior towards Bhir, almost straight east of Ahmadnagar, and, across the Balaghat range, to some­where near Ashti or Kharda, two frontier stations where during the Mahratta wars battles were fought. Marching from there to Barsi and along south by way of Naldurg to Sholapur and thence to Pandarpur. If, as is possible, Sholapur had been the headquarters of Parasuram Deo, the condition that the Bhima was one of the three rivers crossed is not satisfied. But if instead the road taken was from Barsi to Pandarpur, one of the familiar roads, both the Sina and the Bhima would be crossed. Pandarpur was probably the frontier station, and that was the government of Parsudevo Dalvi. We have reference to an inscription of the Hoysala Vīra Sōmēśvara discovered there. It is therefore clear that Pandarpur was the frontier between the two kingdoms of Baḷḷāḷa and Yādava. Therefrom, after holding a council of war, Mālik Kāfūr started at the head of one tuman (a division of ten thousand soldiers), and made a dash upon Dvārasamudra, reaching the place after twelve marches, on the 25th February 1311. Without much of fighting, terms of peace were arranged with Vīra Baḷḷāḷa, not Viranarasinga as stated by Elliot and copied from him by Professor Habib. After the terms of the treaty had been arranged, Mālik Kāfūr probably remained there some time when the main army joined him. After a stay of about twelve days, he left Dvārasamudra on 10th March 1311. Five marches took them to the frontier of Ma’bar. From there two passes had to be crossed, which, according to Professor Habib’s reading, are Tarmali and Tabar. The first according to Elliot’s reading Sarmalai, introduced a certain amount of confusion. Marching through these passes, they came to the bank of a river, which Elliot noted as Kanobari which name Professor Habib says, does not occur in the manuscript. But as a river is mentioned, and, in the next stage of the narrative, it is what is called Kanauri, the river under reference is apparently the Kaveri. There they encamped for the night, and left on the 25th March 1311 towards Birdhul. Unfortunately for us, Amīr Khusrau uses the name Birdhul once for the capital, another time for the ruler, and contributes to make confusion worse confounded. We have to take Birdhul to be the equivalent of Vīra-Chola. That would be all right if it is applied to the capital. But the ruler against whom he went was Vīra Pānḍya, and, in that case, we shall have to take it that the ruler was Bīr or Vīra, ruler of the Chola country for the time being. Amīr Khusrau’s tendency to play upon the word Bir only adds to the confusion. Here the incidents of the war are said to be an attack of Birdhul, Vīra Pānḍya fleeing from there to a place called Kandur on the outskirts of the forests.

Notwithstanding all the destruction that was committed in Kandur, which Amīr Khusrau calls here Kharābābād, which seems to be a term of opprobrium with him and nothing more, the Muhammadan army suspected that Vīra Pānḍya fled to the sea-shore for protection to a fortress of his there, which, according to the reading of Elliot, was given as Jālkoṭa; but Professor Habib’s reading of the manuscript makes it Jāt Kūṭa. Certain people coming from that direction gave the information that Vīra Pānḍya had not gone in that direction; as no information of either the king or of his army was to be had, and as Amīr Khusrau indulges in a pun that Vīra Pānḍya washed his hands off the sea and fled rather to the forests, as offering more satisfactory protection to his army, it is probable Vīra Pānḍya fled towards the Pachamalais on the Salem frontier. The army therefore returned to Kandur, and went forward in search of Vīra Pānḍya through the dense forests surrounding the town. Finding progress impossible they were content to remain at Kandur receiving the submission of the Musalman part of the army left behind. The account of Amīr Khusrau makes it appear, and it may be true as other accounts seem to confirm it more or less, that Mālik Kāfūr was particular about capturing the elephants of the enemy; but so far he was disappointed in not getting possession of as many elephants as he wished to secure, or perhaps as he expected. Just at this juncture information reached him that in a place called Barmatpur—Elliot’s reading Brahmastpur—there was a golden temple and a rich city with temples and idols to be plundered together with some elephants, which they were given to understand were kept there for greater security. The army marched to Barmatpur or Brahmastpur, which, of course, is stated in so many words to be destroyed completely, the images of Śiva and Vishṇu alike being included in the destruction. The whole place seems to have been dug up for buried treasure, and after taking whatever was possible to get, the army was preparing to proceed to the next stage of their work.

On the 11th Zil Qa’d, which seems to correspond to April 1, 1311, they left the Chola country on their march towards Madura. The first stage in the march after four days was a place which the narrative calls Kim here, but which was read Kham by Elliot. Five marches thence they reached Madura, Matra, the ‘dwelling place of the brother of the Rai Sundara Pānḍya.’ The king ‘with his household and all that could be carried, had fled, and the invading army found nothing’ there excepting an empty palace with only three elephants in the temple of Jagannath (Śokkanātha). Mālik Kāfūr’s disappointment was very great. He took possession of the three elephants and sent them all to join the rest of the elephants taken in the course of this invasion; and Amīr Khusrau accounts 512 of such elephants taken so far in this invasion. The anger of the invading general showed itself in destroying by fire both the temple and the palace in the immediate neighbourhood. So far as this account is concerned, the campaign comes to a close here. On Sunday the 4th Zil Ḥijja, which would correspond to the 24th or the 25th of April 1311, the army set forward on the return journey and reached Delhi and presented itself in the durbar of the Sultan on Monday the 29th September 1311, having been five months and two or three days on the return journey. The rest of the account is occupied of course, with the description of the wealth that was carried from the invasion, and what Allāuddīn himself did with the spoils of the war.

Now proceeding to an examination of the route of this invasion, we have already indicated that the road taken this time should have been the high road from Delhi to almost Nagda, from there to Indore, from there to Dhar, across the Narmada to Deogiri; from Deogiri, the army proceeded to Bhir, and thence across on the road to Parenda, from thence southwards to Barsi and across to Pandarpur, the fief of ‘Persdevo’. From there Mālik Kāfūr made a dash at the head of a select and compact body of troops, about ten thousand strong, against Dvārasamudra, Haḷabid of the Hoysalas, under Vīra Baḷḷāḷa. No battle is described, and no battles were obviously fought. But the Baḷḷāḷa submitted and terms of treaty were arranged, sending forward the treaty with the Baḷḷāḷa prince to the headquarters for ratification of the treaty. We must note here that no further campaign is mentioned by Amīr Khusrau, nor is any made of any battle fought, or siege laid in Dvārasamudra. Therefrom five marches took the army to the frontier, by way of the Hoysala country, to that of the Chola-Pānḍya. The dash from Pandarpur upon Dvārasamudra could have been only by one of two well-known routes, either from Pandarpur to Bijapur, and by the eastern road through Anegundi and Hampi straight along as far as Hiriyur in Mysore, and thence to Banawar and Halabid. Returning by the same route as far as Banavar and taking the route to strike the main road at Chikknayakanhalli, the army must have marched forward towards the passes into the Salem District at Hosur; or they must have marched taking the western route through Dvārasamudra and Halabid to Harihar, thence coming down as far as Kadur or even Banavar, and leaving for Dvārasamudra. The main army joined Mālik Kāfūr at Dvārasamudra. From there the road taken was surely the eastern road through Gubbi and Bangalore towards Hosur, Krishnagiri and across the hills in the Salem District. In those days that was a very well-known route, and one of the highways of communication between the country above the ghauts and below. Amīr Khusrau gives us no very particular infor­mation possibly because there was nothing interesting that occurred. The whole route lay through the country of Vīra Baḷḷāḷa, and, once a treaty had been entered into, nothing could well have happened worth mentioning. But from Hosur south, it is a new route and it runs through hostile country. Amīr Khusrau mentions two passes; according to Elliot’s reading, the names are Śarmali and Tabar; but Professor Habib’s reading is Tarmali and Tabar. The objective of the invasion seems to have been to strike the Kaveri some­where. That gives us a little lead where there was actually none. After mentioning the passes the name of the river is given as Kanobari, according to the manuscript of Elliot. But in the manuscript used by Professor Habib, the term Kanobari does not occur; but later on in speaking of their striking camp, the starting point is given as Kanauri, which amounts to as much as mentioning the river before. In both the cases, the river where they came for the night must have been the river Kaveri either on the banks of which, or in the sands of which, they spent the night. The ghat road through Salem leaving Hosur has to pass through Hosur, Krishnagiri, then Dharmapuri on to Salem, Trichengode, Bhavani, across the Kaveri as one route. That is the route taken now-a-days. There is another, Krishnagiri, Dharmapuri, Tāramangalam reaching the banks of the Kaveri higher up than Bhavani perhaps almost opposite the place where now the Meṭṭur dam is being built. Of course, an army marching would certainly see to the facility of crossing the river Kaveri at a place where it can easily be forded. There are two passes here; and whichever of the routes be taken, the army has to pass through Toppur. Toppur is a village where there is a little stream, which is called Toppur river; and as it cuts its way through one of the spurs of the Eastern Ghats, the pass gets the name Toppur pass. It is a well-known place on the road from the plain country into Mysore in days before the railway. There is no need to get through another pass at all if the route through Omalur-Salem be taken; but the mention of a second pass and the definite statement in the account of Amīr Khusrau that there were two passes to cross, give clear indication that the road taken was the other, and probably it is the old ford on the Kaveri at which the river could be crossed with ease to the opposite bank, where in those days there were important towns along which the road ran to the south. Tarmali is probably the Taramangalam river, or the pass a little to the west of Taramangalam. Then after crossing the Kaveri the army left on its march. If they had crossed the Kaveri before marching, which seems to be indicated in the statement of Amīr Khusrau, the road would take them down south as far as Musiri close to the Kaveri, and then the road takes off from the Kaveri into the interior. At Musiri there is even now a well-known ford across the Kaveri. Probably that was a ford even in the thirteenth and fourteenth centuries. Crossing the Kaveri again, although it is not mentioned, the road could be taken on the other side of the Kaveri and the Coleroon to the Chola capital at Jayamkonḍachōḷapuram or Gangaikonḍachōḷapuram, the former of which appears to have been the Pānḍya capital of the Chola country at the time. The route between Musiri and Trichinopoly on the other side of the river would be comparatively inconvenient for an army. There were so many water channels and the country is so thoroughly agricultural that the difficulty would be really great, unless the army had made a detou?? into the interior and marched towards the south to Madura by way o?? Dindigul. The objective being Birdhul, according to Amīr Khusrau, we shall have to fix upon what that is. Undoubtedly that must have been the capital of the Chola country, now the eastern capital of the Pānḍya empire. The capital in those days must have been Jayam­konḍachōḷapuram, which is about six miles to the west of Gangaikonḍa­chōḷapuram on the highroad to Kumbhakonam on the one side, and Trichinopoly on the other on the northern bank of the Coleroon. That was the capital of the great Pānḍya Māravarman Kulaśēkhara, and records of his reign state that he issued his orders from a garden palace outside the city of Jayamkonḍachōḷapuram. But even this does not bring us any nearer to Birdhul. Since the phonetic approach that we could make to Birdhul is Vīra-Chōḷa, and, as we hear often of halls and palaces bearing names of these princes in some of these cities, such as for instance Muḍikonḍaśoan in Gangaikonḍachōḷapuram, it is just possible that this city Jayamkonḍachōḷapuram contained halls which may have been called Vīraśoan. Vīraśoan was a common name of a large number of princes, and there were two emperors who bore the name, Vīrarājēndra ruling from A.D. 1063 to A.D. 1070, and Kulot­tunga III not long before the date of the invasion, who had the title Tirubhuvana Vīra. Other indications point closely to that as the centre of the attack. It is not very far from Chidambaram, where often, these princes anointed themselves in the hall of the holy place there, soon after the royal ceremony in the capital. If then Jayamkonḍa­chōḷapuram is what is to be taken to be the Birdhul, what is Kandur and the forests near about? Kandur is undoubtedly Kaṇṇanūr on the northern bank of the Coleroon about six miles from Śrīrangam across the comparatively big river. Kaṇṇanūr must have been a place of importance at the time, as that happened to be the Hoysala capital down to the time of Vīra Baḷḷāḷa. It was only some years after Vīra Baḷḷāḷa began to rule that the whole of the Hoysala territory above the Ghats and below united into one with Vīra Baḷḷāḷa as ruler. Till then his uncle, and after him that uncle’s son ruled over the country south with their principal capital at Kaṇṇanūr with an alternative which finds mention as Kundāṇi, which is now satisfactorily located in the Baramahals of the Salem District, not very far from Hosur. This Kaṇṇanūr might have been on the way of march towards Vīraśoan, but if the invasion went in the first instance to the capital of the country, they might have passed Kaṇṇanūr by. Kaṇṇanūr would otherwise answer to the description given in Amīr Khusrau. It is on the outskirts of the forest almost on three sides except on the side towards Srīrangam. It probably was the centre of a Muhammadan population even before the days of Mālik Kāfūr, as there were a number of places in which Muhammadans were settled for purposes of commerce, at any rate, in the interior, as there were very old settlements of these in a number of places along the coast. Vīra Pānḍya when he was attacked in Birdhul retired after a faint resistance to Kandur, where the Islamic army chased him. But when they went to Kandur, Kharābābād as it is called, they did not find Vīra Pānḍya or his army there. They inferred naturally that he must have gone away in the other direction to another fortress of strength, where he could find efficient protection. According to Amīr Khusrau’s account, it must be a place towards the sea, the sea itself contributing to its defence. According to Elliot’s manuscript, it is Jālkoṭa, but Professor Habib reads it Jāt Kuṭa. There is not much substantial difference in the latter part of the word. Koṭa and Kūta could be easily mistaken in pronunciation. But whether it is Jāt or Jāl would make a substantial difference in sense. With the reading Jāt, it is rather hard to make anything out of it. Jāl, water, may make some sense, and the way that Amīr Khusrau seems to play upon the term sea and the statement that Bir, Vīra Pānḍya washed his hands off the sea, would seem to indicate that perhaps the correcter reading would be Jāl. In such a case, the place would be Jālkōṭa or water fortress, as Amīr Khusrau is actually describing Vīra Pānḍya as having run away for the protection of the sea. This must have been the strong fortress, which, in the centuries following, played an important part, and which in the days of the British used to be called Devacotta, really Tīvukōṭa, the fortress on the island, popularly spoken of as Tīkoṭai, which is at the mouth of the Coleroon, but in those days extending perhaps northwards to a considerable distance into the island which the river has formed there. There used to be a flourishing seaport at the mouth of the Coleroon, which in the days of the Cholas used to be called Jayaṁkonḍa Chōḷa­paṭṭinam, or briefly Jayangonḍapaṭṭinam. But the Muhammadan army did not march there having had information that Vīra Pānḍya had not gone in that direction. Their attempt to follow him into the forest proved infeasible, and they had no alternative but to content themselves with what they had been able to do. Perhaps the inevitable elephants had not been secured in number. Timely information came to Mālik Kāfūr that the elephants of Bir were kept in a strong place which Professor Habib reads as Bharmatpur, and which Elliot read as Brahmastpur. Either of the readings would be equivalent of the Hindu Brahmapuri, which I have indicated from the circumstantial account of Amīr Khusrau himself to stand for Brahmapuri-Chidam­baram, which according to the account of Amīr Khusrau, Mālik Kāfūr sacked. He destroyed the temples and the idols there, and even dug up the place for buried wealth, having secured some elephants also. He broke camp from this locality and started on the invasion of Madura, Matra, as Amïr Khusrau calls it, and Mardi as Wassaf writes it, more truly catching up the popular name of the town. The first stage in the march which seems to have taken them about four days was mentioned as Kham by Elliot, and Kim by Professor Habib. Either way it does not take us nearer to a satisfactory identification, which may be regarded as certain. We have to remain content with what I stated in my South India and Her Muhammadan Invaders on pages 104-5, that probably it stands for Kaḍambavanam, which would be about sixty miles from Chidambaram. It may have taken five marches to reach it. Another five marches from there, they reached the town of Madura, the habitual capital of the Pānḍyas, even under Vīra Pānḍya. The capital had been evacuated by the ruler who carried away his treasure and household to a place which is mentioned as Mankul, which Elliot attempted to identify with Namakkal. Namakkal is too far out for the purpose. It is there probably the two Mangalams, Mēla (Upper) and Kīa (Lower) on the Western Ghats, which may be regarded as places of security. Not finding anything but three elephants, the Muhammadans set fire to the temple itself and taking the three elephants returned making up a total of 512 elephants for this invasion. This is as far as Amīr Khusrau’s account according to the Khaẓā’inul Fuṭūḥ or Tarik-i Alāi takes us.

Before we close this introduction we ought to refer to two minor points which are neither of them mentioned in this account. Firishta is responsible for the statement, given in the portion translated by Professor Habib himself, that Mālik Kāfūr, in the course of his southern invasions, constructed a masjid. It is called Masjid-i-Alāi, Allāuddin’s Mosque, at ‘Sit Band Ramisar’, and adds that the mosque was to be seen existing in his days early in the seventeenth century. The designation Sita Band Ramisar would lead one readily to take it that it is Sētubanda Rāmēśvara, or Ramēśwaram, where a big dam across the sea ascribed to Rama exists. But Firishta’s own descrip­tion gives the lie direct. He says that the Malik Naib after over­coming Bikal Dev, the Raja of the Carnatic, plundered the country, and it is in that invasion and in that connection that he says that he built a mosque in that country of Carnatik, and later, in the same connection, that the mosque could be seen in that country and describes it as ‘the port of Dur Samandar on the shore of the sea of Ummam’. Sea of Ummam is certainly the Arabian Sea, and the port or Dur Samandar must be a port in the country of Dvārasamudra, that is the Hoyśala country. It cannot therefore refer to Ramēśwaram by any stretch of language. There is no reference to such a thing in the account of Amīr Khusrau, or in the abridged account of Zia Barni. It is just possible that a raid was undertaken towards the west coast. But this has nothing whatever to do with Ramēśwaram in the distant south. That would lead us to the question whether Ramēśwaram, or that region, was ever at all invaded by Mālik Kāfūr. We have seen that Mālik Kāfūr, in the course of these wars, was in the habit of making dashing raids against various places. From the capital of Hoysala a raid to the west coast may seem possible. It is from Pandarpur that he first undertook a raid towards Dvārasamudra. It will be noticed that from the camp of Bas‘īrāgarh, a raid was undertaken against Sarbar on the occasion of the invasion of Warangal. It is not at all unlikely that he sent out a raid, or led it himself from his camp at Madura towards Ramēśwaram as a likely place where the elephants that he wanted so much and the wealth of the Pānḍya might lie hidden. In another work of Amīr Khusrau called Āshika, he speaks of an invasion up to the shores of the sea of Lanka, against the ruler whom he called Pānḍya Guru. He mentioned his capital by the name Faṭan where there was an idol temple. This, in all probability, is no other than Ramēśvarapaṭṭinam, as we may call it now, as well as Peria­paṭṭinam and another Paṭṭinam on the opposite coast of India. Excepting for this raid we have no information whatever that Mālik Kāfūr had anything more to do with Ramēśwaram. It may therefore be taken to be that he carried his raids as far as Ramēśwaram. It is hardly possible that he built a mosque there though there might have been one already, and that Firishta’s account is the result of a confusion that the mosque was built at Sētu-Rāmēśwaram when he speaks of the sea of Omman. For a further discussion of this, reference may be made to South India and Her Muhammadan Invaders.

We must close the introduction with pointing out the services that Professor Habib has rendered to students of Indian History by giving such a good translation of this difficult work of Amīr Khusrau, and adding to the translation itself material from other works bearing on the subject, which would enable a student to arrive at a satisfactory conclusion regarding the invasions of the south under Allāuddin. This work needed the doing and we are grateful indeed to Professor Habib for having taken up this work and done it so well.