There is but one chamber, this has been divided by a thin brick wall by some Mahomedan fakir, perhaps several centuries back, the doorway or aperture to which is so small as to have prevented my entering it, but I felt the end with a 10 foot rod. From the fragments found scattered, I conclude that there must have been a very handsome temple here of very early date.

Retracing our steps and turning to the south, we descend through a passage between a row of huge blocks of sienite, which had formed part of the southern barrier to the great enclosure, we then come on to the plain, then turning to the left (or east) and continuing to a tope of Tar-trees under the Nag-arjuni's frightful crags, we reach a flight of steps, about the centre of the hill; after an ascent of about 60 feet a narrow terrace is found continued along the side of the rock, in the centre of which an Egyptian doorway leads you into a splendid ovalshaped and vaulted room, polished in the usual manner; over the doorway is a square polished surface containing the inscription No. 1, pl. IX, is 3 of Prinsep's plate, and on the left hand side of the passage or thickness of the rock, is that given in his plate XXXIV. and translated at page 673; on the opposite side are some more recent scrawls.

This cave is inhabited and has been so for many years, by Mahomedan saints; there is a small mosque before the door, more than a century old; the cave is called Nag-arjuni, whether from the Budhist saint of that name having lived there, or from mere fancy it is not possible to decide, though as the name Sidheswar has been preserved, and if my version of" Satgurba," be correct, I see no reason to doubt the inscription, moreover I am inclined to think that it is the very peraputetic chamber named in the Pâli annals in which Annund Múni performed his austerities.

Having described the caves I must conclude with Dhuravat.

I have already said that it is the north-westernmost end of the cluster of hills. I visited this place by moonlight, therefore had not so good an opportunity of examining the locality, however, I saw sufficient to enable me to decide that there is the site of a Budhist temple. On the lowest hillock, at the head of which is a fine tank called Chundoke, many idols and miniature Chaityas, such as are found all over the district, are placed in and about a modern temple to "Nirsinha" on the east bank of the tank; there is one very remarkable figure of a

man with twelve arms, each hand holding a lotus; it is a Budhist sculpture. I was told of several other figures in the vicinity, but had no leisure to examine them, those I saw were comparatively modern, mostly well executed. To the northward of the tank is a high mound of bricks and rubbish, perhaps the ruins of a monastery or of some of the buildings of the ancient town, of which nothing else remains.

I will now offer a few words on the inscriptions, of which there are in all 29; including that on the idol at Sidheswar, six are in the old Pâli, three in the Gupta, and three in an unknown character, to which I shall invite particular attention, and the remainder are in various types of Nagree, from the earliest to latest date.

Plate IX. Nos. 1 and 2, are those numbered 3 and 2, in Prinsep's plate. No. 3 had hitherto been overlooked, being in the same cave as the long inscription No. 9, plate X. It will be perceived that there is a slight difference in some of the words of the three, perhaps errors in cutting, otherwise they are verbatim. The same, excepting the initial name (of the cave), I have neither books to refer to nor pundits to consult by which I might explain these variations, therefore I must content myself with inviting the attention of those who are more fortunate, and who are better scholars.

In fig. 1 the word

"Gopi" is clear enough, but instead of the last word of the inscription being ↓ Aliyam, it seems to be

↓↓↓· Sooliyam, though I am inclined to think it is merely a mistake of the engraver. LLOJ “Nisiti" is written LCO- “Nisita” both in Nos. 1. and 3., No. 2. on the contrary has neither change; indeed with the exception of five letters, purposely hammered out, it is quite perfect, (a stronger proof of the soundness of Prinsep's conjectures could not be needed, the copy he had being very imperfect ;) however, knowing what they should be, it was no difficult matter to trace them, but it must be observed that all the inscriptions in the lath or Pâli character have had the letters ground and polished after cutting, to which circumstance their better preservation must be attributed; moreover all have been cut on a polished surface. I speak of those I have myself seen; those in our museum afford proofs.

In No. 2. the word

is deserving of notice, the second letter

being more like h. but the mark is placed lower down, and may be equivalent to the short ikar of the Sanscrit, though it has been supposed not to exist in the Pâli; this would at once make it of anta. aЯte appears to have no meaning.

instead

No. 3 over the doorway of the cave which appears to have escaped notice, has the word as the meaning of which I know not; in other respects the inscription is a repetition of those before named.

We now come to the three remaining Pâli writings that have hitherto been overlooked. The first (figure 4) is the most perfect, though the five last letters which I feel warranted in restoring have been hammered out as before described. Upon refering to Prinsep's papers on the pillar inscriptions P. 471, Vol. VII. I find that the sentence " Duva dus vasa bhisiténa" ↓↓↓ is common to them. I am enabled therefore to read that "by the beloved Rajah in the 12th year of his reign, this cave was caused to be excavated, &c." the remainder I cannot render for reasons above given. Thus much would seem to point to the same person as author both of the pillars and of these caves, and if the similarity of design and execution be considered a criterion, we may infer that it was Dusarat himself whose name is repeated with the title "beloved of the gods" in the three first named inscriptions. This reasoning it will be seen throws doubt on the assumption that Asoka was the author, a doubt Prinsep himself always entertained. Indeed, if the conjectures I have made on reading the passage before quoted of the Pâli annals, are correct, there are none, that he was not.

If we may judge by the unfinished state of caves, (Nos. 4 and 7, plate VIII.) we shall naturally conclude that they are of later date than those bearing Dusarut's name; one difficulty would thereby be removed were it not that this prince (if Prinsep be right), was the third in descent from Asoka; but it suggests a further conjecture, i. e. whether this Dusarut may not have been the very deified personage of the purans, king of Ajudhia and father of the hero Ram, whose history I believe to be a mythological tale of a real event. The conquest of Lunka or Ceylon by him may be in reality that made by a prince, whose title was Devanam Prya, or Devanam Prya Dasa, for either title conveys the same meaning. I throw out these as hints for the consideration of those who have read more and have better opportunities for study than myself.

I may here further digress and allude to the rude Budha sculptures I found at Bôdh Gyah and which I am about to lay* before the society. In these, buildings are represented with arched entrances precisely of the design of that of the great cave, which again (like those in Cuttack) are miniatures of similar designs in the caves of Carli and others in the west of India. Now it is common for Archæologists to found their arguments as to the age of ancient buildings, upon comparing them with the paintings which illumine ancient manuscripts of known date, it being evident that such forms must have been in existence at, or prior to, the representation being made, the same rule must apply to sculptures such as those of Bôdh Gyah, therefore if they belonged to Asoka's great temple, the works they represented existed before it, or at the same time at least; but as these sculptures represent buildings on rocks, we may incline to the first opinion. That the sculptures belonged to the same period as the first pillars there can be scarcely a doubt, for they are of precisely the same stone, and one the quarries of which have not been discovered; this is of itself a remarkable circumstance; besides, they bear inscriptions in the identical character, and record gifts in the same style as those on the posts of Bhilsa.

I now return to my inscriptions. Fig. 5 differs somewhat in the reading, there being an adjective between the opening sentence, and the word or cave. This word after the most careful examination appears to be

"Nigôp," the "gôp" is clear enough, which is curious, for it is at the entrance of the very cave which I have described as having a Dagôp or Chaitya, the remaining obliterated letters amounting in number to that required for such, together with parts of them still visible, admit of our supposing the sentence to have been the same as the others, i. e. " for the one of Budhist ascetics."

No. 6 is too imperfect to be made out without the aid of a clever pundit, and needs such a Pâli scholar as Rutna Pala who assisted Prinsep. I must content myself with inviting the attention of others possessing greater advantages. The mark will not escape notice, it is found on coins, and in the inscriptions of the Saindharee caves and that of Kund

* See notes on Sculptures at Bodh Gyah, p. 334 of the present Volume.

girri in Cuttack. This inscription is immediately over the left corner of the entrance, but so weatherworn and mutilated that a casual observer would not perceive it. A sentence seems to have existed over the door, but is now become eligible.

*

No. 7 is that published as 15 by Prinsep (see pl. XXXV. Vol VI.) and from the awkward manner in which the separate sheets of impressions were taken by the Moonshee, gave so much trouble, and rendered the reading doubtful; by the impression I now submit, it will be seen that, instead of its occupying three distinct spaces, the whole is comprised in one, and, as conjectured by Prinsep, it fills the spandril or space between the arched head and the top of the square doorway. I believe Kamulakanta to have been right when he pronounced that the first two lines had no connection with the four last. This struck me at first sight. I have since read the remarks, the letters are smaller, and I think have been added at a somewhat subsequent date; there can be no doubt that both have been an afterwork, perhaps centuries later than the

caves.

No. 8 has also been rendered by Prinsep in the same volume, but it may be as well to compare the present accurate impression with the former; it is engraved within the jaumb of the doorway to the Nagarjuni (oval) cave, the edges are rough owing to the want of skill in cutting.

No. 9, plate X. appears hitherto to have (together with No. 3, (before described), escaped notice. I trust that some scholar will come forward to translate it, should I not be able with the assistance of a clever pundit to do so, but I shall first beg to invite Saroda-purshad to undertake the task. I feel sure it could not be entrusted to better hands. The character is the same as that of the two foregoing numbers, the dates, therefore, may not much differ. This is likewise cut within the jaumb of the small cave, fig. 3, plate VIII.

No. 10 is inscribed on either side of the head of a female figure or idol on the Sidheswar temple: It is a very rude performance.

Nos. 11, 12, 13, and 14, or figs. 7, 8, 9, plate IX. are the curious characters to which I would invite the attention of our French and German fellow-labourers. I have remarked the same characters on the Allahabad pillar, and in the caves of Cuttack. Prinsep refers to the * This refers to the inscriptions exhibited at Capt, Kittoe's lecture. -Eps.