facts I conclude, with a probability amounting almost to certainty, that the temple from which these pillars were obtained consisted of 20 columns only. On No. 12 shaft there is the word Kachal in Nagari letters on one face, with the date of 1124 on another face, which, referred to the Vikramâditya Samvat, is equivalent to A. D. 1067, at which time Anang Pâl II., the founder of Lâlkot, was reigning in Dilli.

But the mason's marks on the stones of this temple were not confined to the pillars, as I discovered them on no less than 13 different portions of its entablature. These marks are more than usually detailed; but, unfortunately, in spite of their length and apparent clearness, I am still unable to make them out completely.*

There is a peculiarity about the numbers of the pillars which is worthy of note. Each cypher is preceded by the initial letter of the word for that number. Thus, 3 is preceded by ti for tin, 10 by da for das, and 16 by so for solah. The same style of marking would appear to have been used for a second temple, as I found a pillar of another pattern with the number du 2, and a pilaster of the same kind with

*See Plate XXXVII. for copies of these mason's marks, and a drawing of one of the pillars. During a visit of a few hours in the present year, 1871, I found two numbered pillars of a different kind, with the Nos. 2 and 19, showing that a second temple, destroyed by the Muhammadans, must have been supported on not less than 20 pillars. I found also a mason's record of five lines on a third variety of pillar, but the letters are faint and difficult to read. I can make out a notice of 7+6+5+8, or 26 pillars altogether, of which I discovered 6 in the cloisters.

i 19.* Sixteen bases of the first pillar have recessed angles, and four are plain squares. In this case the temple would have had 4 pillars (probably an outer row) of one pattern, and 16 of another kind, but all of the same height.

The dimensions of these inscribed pillars are as follows:

Ft. In. Ft. In.

The only other Hindu remains are the two forts of Lalkot and Rai Pithora, which together formed the old Dilli of the Musalmâns, after the building of a new fort of Siri by Ala-ud-din Khilji. Of these two, the older fort of Lalkot has hitherto remained unknown, being always described by Musalmâns as a part of the fort of Rai Pithora. It is called Siri by Lieutenant Burgess, who made a survey of the ruins of Dilli in 1849-50, and the same name is given to it by Messrs. Cope and Lewis in their interesting account of Firuzabad, published in the Journal of the Archæological Society of Delhi for 1850. The reasons which induce me to identify this fort with the Lálkot of Anang Pâl have already been given when speaking of the re-founding of Dilli, and the reasons which compel me to reject its identification with Siri will be detailed when I come to speak of that place.

The Fort of Lálkot, which was built by Anang Pâl in A. D. 1060, is of an irregular rounded oblong form, 24 miles in circumference. Its walls are as lofty and as massive as those of Tughlakabad, although the blocks of stone are not

These two pillars are 4 feet 10 inches high, and 11 inches square. I found 13 pillars of almost the same pattern, but of somewhat large dimensions, being 5 feet 34 inches high, and 133 inches square. The commonest pillar is of a similar pattern, but with the addition of human figures on the lower faces of the shaft, and a deep recessed ornament at the top of the shaft. Of this kind I counted 78 pillars during my last visit in the present year

1871.

so colossal. By different measurements I found the ramparts to be from 28 to 30 feet in thickness, of which the parapet is just one-half. The same thickness of parapet is also derived from the measurement given by Ibn Batuta in A. D. 1340, who says that the walls were eleven cubits thick. Accepting this measure as the same that was in use in Firuz Shah's time, namely, of 16 inches, as derived from the length of Firuz Shah's pillar, the thickness of the walls of old Dilli was 14 feet. These massive ramparts have a general height of 60 feet above the bottom of the ditch, which still exists in very fair order all round the fort, except on the south side, where there is a deep and extensive hollow that was most probably once filled with water. About one-half of the main walls are still standing as firm and as solid as when they were first built. At all the salient points there are large bastions from 60 to 100 feet in diameter. Two of the largest of these, which are on the north side, are called the Fateh Burj and the Sohan Búrj. The long lines of wall between these bastions are broken by numbers of smaller towers well splayed out at the base, and 45 feet in diameter at top, with curtains of 80 feet between them. Along the base of these towers, which are still 30 feet in height, there is an outer line of wall forming a raoni or faussebraie, which is also 30 feet in height. The parapet of this wall has entirely disappeared, and the wall itself is so much broken, as to afford an easy descent into the ditch in many places. The upper portion of the counterscrap walls has all nearly fallen down, excepting on the north-west side, where there is a double line of works strengthened by detached bastions.

The positions of three of the gateways in the west half of the fort are easily recognized, but the walls of the eastern half are so much broken that it is now only possible to guess at the probable position of one other gate. The north gate is judiciously placed in the re-entering angle close to the Fateh Bûrj, where it still forms a deep gap in the lofty mass of rampart, by which the cowherds enter with their cattle. The west gate is the only one of which any portion of the walls now remains. It is said to have been called the Ranjit gate. This gate-way was 17 feet wide, and there is still standing on the left hand side a large upright stone, with a grove for guiding the assent and descent of a portcullis. This stone is 7 feet in height above the

rubbish, but it is probably not less than 12 or 15 feet. It is 2 feet 1 inch broad and 1 foot 3 inches thick. The approach to this gate is guarded by no less than three small out-works. The south gate is in the southmost angle near Adham Khan's tomb. It is now a mere gap in the mass of rampart. On the south-east side there must, I think, have been a gate near Sir Thomas Metcalfe's house, leading towards Tughlakabad and Mathura.*

Syad Ahmad states, on the authority of Zia Barni, that the west gate of Rai Pithora's Fort was called the Ghazni Gate after the Musalmân conquest, because the Ghazni troops had gained the fortress by that entrance. I feel satisfied that this must be the Ranjit Gate of Lâlkot for the following

reasons:

1st.-The Musalmâns never make any mention of Lâlkot, but always include it as a part of Rai Pithora's Fort.

2nd. The possession of the larger and weaker fortress of Rai Pithora could not be called the conquest of Delhi, while the stronger citadel of Lâlkot still held out.

3rd. The evident care with which the approach to the Ranjit Gate has been strengthened by a double line of works, and by three separate out-works immediately in front of the gateway itself, shows that this must have been considered as the weakest point of the fortress, and therefore that it was the most likely to have been attacked. For this reason I conclude that the Ranjit gate was the one by which the Musalmans entered Lâlkot, the citadel of Dilli, and that, having proved its weakness by their own success, they at once proceeded to strengthen the works at this point for their own security. A case exactly similar occurred less than forty years afterwards, when the Emperor Altamsh, having gained an entrance into the fortress of Gwalior by the deep ravine on the west side called Urwahi, immediately closed it by a massive wall, to prevent his enemies from taking advantage of the same weak point. I believe that the western

See Plate No. XXXVI. for an enlarged plan of Lâlkot, showing the positions of the different gates. It seems probable that the western half of Lalkot was once cut off from the eastern half, as there are traces of walls and ramparts running from the Sohan Bûrj on the north direct south towards Adham Khan's tomb. I traced these walls as far as the ruined building to the west of Anang Pal's tank. The western portion would have been the citadel of Lâlkot under Anang Pål, before the accession of Rai Pithora. My Assistant, Mr. J. D. Beglar, has discovered a gateway in the southern half of this wall, between Adham Khan's Tomb and the Jog Maya temple.

gate was called the Ghazni Gate for the simple reason only that Ghazni lies to the west of Delhi.

The Fort of Rai Pithora, which surrounds the citadel of Lâlkot on three sides, would appear to have been built to protect the Hindu city of Dilli from the attacks of the Musalmâns. As early as A. D. 1100, the descendants of Mahmud, retiring from Ghazni before the rising power of the Saljukis, had fixed their new capital at Lahor, although Ghazni still belonged to their kingdom, and was occasionally the seat of Government. But a new and more formidable enemy soon appeared, when the celebrated Muäz-uddin Sâm, commonly called Muhammad Ghori, after capturing the cities of Multan and Parshâwar, appeared before Lahor in A. D. 1180, and put an end to the Ghaznavide dynasty by the capture of their capital in A. D. 1186. The danger was now imminent, and only a few years later we find the Ghori King in full march on Ajmer. But the Raja of Dilli was well prepared for this invasion, and, with the aid of his allies, he defeated the Musalmâns with great slaughter at Tilaori, midway between Karnâl and Thanesar. As the first appearance of the formidable Ghoris before Lahor corresponds so nearly with the accession of Prithivi Raja, I think it very probable that the fortification of the city of Dilli was forced upon the Raja by a well-grounded apprehension that Dilli itself might soon be attacked; and so it happened, for within two years after the battle of Tilaori the Raja was a prisoner, and Dilli was in the possession of the Musalmâns.

The circuit of Rai Pithora's Fort is 4 miles and 3 furlongs, or just three times as much as that of Lálkot. But the defences of the city are in every way inferior to those of the citadel. The walls are only half the height, and the towers are placed at much longer intervals. The wall of the city is carried from the north bastion of Lâlkot, called Fateh Búrj, to the north-east for three-quarters of a mile, where it turns to the south-east for 11⁄2 mile to the Damdama Búrj. From this bastion the direction of the wall for about one mile is south-west, and then north-west for a short distance to the south end of the hill on which Azim Khan's tomb is situated. Beyond this point the wall can be traced for some distance to the north along the ridge which was most probably connected with the south-east corner of Lâlkot, somewhere in the neighbourhood of Sir T. Metcalfe's house.